{
  "schema": "verified-recent-results.v1",
  "generated_at_utc": "2026-10-14T11:08:01.350688+00:00",
  "scope_note": "Completed, verified attempts only. Running and unrun tasks are omitted, not failures.",
  "suites": [
    {
      "id": "deepswe-primary-20261012",
      "label": "DeepSWE \u00b7 primary \u00b7 2026-10-12",
      "adapter": "deepswe",
      "benchmark": "DeepSWE v1.1",
      "campaign": "tonight-20261012",
      "native_score_semantics": "feature-to-patch and patch-to-patch verifier checks",
      "tasks": [
        {
          "task_id": "actionlint-action-pinning-lint",
          "official_prompt": {
            "title": "Add action pinning linting for actions and reusable workflows",
            "description": "Add a configurable lint rule that enforces pinned versions for action and reusable workflow references.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/actionlint-action-pinning-lint",
            "instruction_sha256": "02472c08b7de7ca4d7d709a3b53db835fe8bcc76a887a13c18dc5aba86bd37f9",
            "task_toml_sha256": "a171ddc6849f81f146c05b5fa230341fbc691e26b90edb4d3d80efd274141f66",
            "language": "go",
            "category": "feature_request"
          },
          "repository": "https://github.com/rhysd/actionlint",
          "base_revision": "0bdc95715fa58f64e3fd6e63b0f89be8733cbbab",
          "runs": [
            {
              "run_id": "tonight-20261012/actionlint-action-pinning-lint__8JiBsSU",
              "task_id": "actionlint-action-pinning-lint",
              "campaign": "tonight-20261012",
              "outcome": "pass",
              "score": {
                "feature": {
                  "passed": 55,
                  "total": 55
                },
                "preservation": {
                  "passed": 145,
                  "total": 145
                },
                "reward": 1
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "agent": "OpenCode/Pier",
                "runtime": "Strata 0.1.42",
                "agent_timeout_seconds": 2700
              },
              "timing_seconds": {
                "agent_execution": 955.0015,
                "environment_setup": 25.700454,
                "verifier": 26.950975
              },
              "usage": {
                "input_tokens": 7889324,
                "cached_tokens": 7825301,
                "output_tokens": 45248,
                "peak_context_tokens": 108867,
                "steps": 102,
                "requests": 103
              },
              "patch": {
                "bytes": 52001,
                "sha256": "f6fb86b28d3fc159e8f9ed295e82783dbbfd2cf81efe1f6c27fedd07af008f77"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [
            {
              "model": "gpt-6-astra",
              "provider": "openai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_gpt_6_astra_xhigh",
              "reasoning_effort": "xhigh",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 2,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 3,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 4,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/actionlint-action-pinning-lint",
              "source_html_sha256": "93be06c94587036cdb888b89bb283bc46f6027074281094bae1ba0a3302f2081"
            },
            {
              "model": "claude-opus-5",
              "provider": "vertex_ai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_claude_opus_5_max",
              "reasoning_effort": "max",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/actionlint-action-pinning-lint",
              "source_html_sha256": "93be06c94587036cdb888b89bb283bc46f6027074281094bae1ba0a3302f2081"
            }
          ],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "fastapi-deprecation-response-headers",
          "official_prompt": {
            "title": "Add deprecation, sunset, and successor headers to FastAPI routes",
            "description": "Add runtime Deprecation, Sunset, and Link headers plus OpenAPI metadata and tracking middleware for deprecated routes.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/fastapi-deprecation-response-headers",
            "instruction_sha256": "b8e0cb43fe2850b2173c8b5e940d463345b96aed6a2b3e7c0c559deda468c568",
            "task_toml_sha256": "b00b517b6f66358b735c0764dfed3b92a69079e366bb3ee5c6c2a9c6bf1b75c8",
            "language": "python",
            "category": "feature_request"
          },
          "repository": "https://github.com/fastapi/fastapi",
          "base_revision": "11614be9021aa4ac078d4d0693a8b5250a1010d8",
          "runs": [
            {
              "run_id": "tonight-20261012/fastapi-deprecation-response-hea__qUmcQMZ",
              "task_id": "fastapi-deprecation-response-headers",
              "campaign": "tonight-20261012",
              "outcome": "graded-model-failure",
              "score": {
                "feature": {
                  "passed": 136,
                  "total": 137
                },
                "preservation": {
                  "passed": 3134,
                  "total": 3134
                },
                "reward": 0
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "agent": "OpenCode/Pier",
                "runtime": "Strata 0.1.42",
                "agent_timeout_seconds": 2700
              },
              "timing_seconds": {
                "agent_execution": 693.513478,
                "environment_setup": 26.03664,
                "verifier": 60.93147
              },
              "usage": {
                "input_tokens": 2222044,
                "cached_tokens": 2191113,
                "output_tokens": 32615,
                "peak_context_tokens": 63202,
                "steps": 50,
                "requests": 51
              },
              "patch": {
                "bytes": 62900,
                "sha256": "54f984444192b80135bf13a922da7c6a23aa482b1f1bd997fcb0b443ab74c10d"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [
            {
              "model": "gpt-6-astra",
              "provider": "openai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_gpt_6_astra_xhigh",
              "reasoning_effort": "xhigh",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 2,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 3,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 4,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/fastapi-deprecation-response-headers",
              "source_html_sha256": "a51b3d13dd8efd3077af4bcd9cc6a59caf286e2d1a29fe5fd6a63cf405021f21"
            },
            {
              "model": "claude-opus-5",
              "provider": "vertex_ai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_claude_opus_5_max",
              "reasoning_effort": "max",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/fastapi-deprecation-response-headers",
              "source_html_sha256": "a51b3d13dd8efd3077af4bcd9cc6a59caf286e2d1a29fe5fd6a63cf405021f21"
            }
          ],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "expr-try-catch-errors",
          "official_prompt": {
            "title": "Add try/catch error recovery to expr",
            "description": "Add expression and block-level error recovery with try, catch, finally, throw, retry, and errtype.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/expr-try-catch-errors",
            "instruction_sha256": "6f19f8363f75963c69072f03aa7515ffa304d6b726f85b09188760ae682e3447",
            "task_toml_sha256": "1206c54353bb7982f44be6473f2bc2837fb2826cc674d1a672612648160ede54",
            "language": "go",
            "category": "feature_request"
          },
          "repository": "https://github.com/expr-lang/expr",
          "base_revision": "851b241a301f7c74646e65e4009c69cf290993a8",
          "runs": [
            {
              "run_id": "tonight-20261012/expr-try-catch-errors__GfLR6AT",
              "task_id": "expr-try-catch-errors",
              "campaign": "tonight-20261012",
              "outcome": "graded-model-failure",
              "score": {
                "feature": {
                  "passed": 73,
                  "total": 79
                },
                "preservation": {
                  "passed": 66265,
                  "total": 66265
                },
                "reward": 0
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "agent": "OpenCode/Pier",
                "runtime": "Strata 0.1.42",
                "agent_timeout_seconds": 2700
              },
              "timing_seconds": {
                "agent_execution": 1599.637729,
                "environment_setup": 25.077473,
                "verifier": 52.316678
              },
              "usage": {
                "input_tokens": 10770815,
                "cached_tokens": 10555924,
                "output_tokens": 69440,
                "peak_context_tokens": 150995,
                "steps": 106,
                "requests": 107
              },
              "patch": {
                "bytes": 32070,
                "sha256": "9c9fed859653f4605a84148603f747b18053bec95266bedf4aa633a8d1d85a90"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [
            {
              "model": "gpt-6-astra",
              "provider": "openai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_gpt_6_astra_xhigh",
              "reasoning_effort": "xhigh",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 2,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 3,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 4,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/expr-try-catch-errors",
              "source_html_sha256": "c2f1f329e75cc26b8b6fa87fe24bf06dd4d5ad37f29914f13bcb18c8a0468924"
            },
            {
              "model": "claude-opus-5",
              "provider": "vertex_ai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_claude_opus_5_max",
              "reasoning_effort": "max",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/expr-try-catch-errors",
              "source_html_sha256": "c2f1f329e75cc26b8b6fa87fe24bf06dd4d5ad37f29914f13bcb18c8a0468924"
            }
          ],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "drizzle-orm-window-function-builders",
          "official_prompt": {
            "title": "Add typed window function builders with OVER clauses",
            "description": "Add typed query-builder helpers for SQL window functions, named windows, and frame specs.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/drizzle-orm-window-function-builders",
            "instruction_sha256": "72fc9abfb71ae51619dd571f8034beb55bfd8a86741c901560d3ff7a905032a8",
            "task_toml_sha256": "6e1b4327782244e00702f20bd33921a43dbfb7b61613197a6f4f5e839bbc349b",
            "language": "typescript",
            "category": "feature_request"
          },
          "repository": "https://github.com/drizzle-team/drizzle-orm",
          "base_revision": "e8e6edfef5ca69c6188d320388ad440265911057",
          "runs": [
            {
              "run_id": "tonight-20261012/drizzle-orm-window-function-buil__xrssWis",
              "task_id": "drizzle-orm-window-function-builders",
              "campaign": "tonight-20261012",
              "outcome": "graded-model-failure",
              "score": {
                "feature": {
                  "passed": 121,
                  "total": 130
                },
                "preservation": {
                  "passed": 566,
                  "total": 566
                },
                "reward": 0
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "agent": "OpenCode/Pier",
                "runtime": "Strata 0.1.42",
                "agent_timeout_seconds": 2700
              },
              "timing_seconds": {
                "agent_execution": 936.194765,
                "environment_setup": 42.774546,
                "verifier": 24.028978
              },
              "usage": {
                "input_tokens": 4071502,
                "cached_tokens": 4031337,
                "output_tokens": 43709,
                "peak_context_tokens": 83613,
                "steps": 88,
                "requests": 89
              },
              "patch": {
                "bytes": 42441,
                "sha256": "bde06a9f1513446f378232041031dfed4a0ddc314007fff781a4618060eb1204"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [
            {
              "model": "gpt-6-astra",
              "provider": "openai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_gpt_6_astra_xhigh",
              "reasoning_effort": "xhigh",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 2,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 3,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 4,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/drizzle-orm-window-function-builders",
              "source_html_sha256": "d9285c39d82339465a6b8a6f99512f570092653f779e1f3e2ba8976c84acc346"
            },
            {
              "model": "claude-opus-5",
              "provider": "vertex_ai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_claude_opus_5_max",
              "reasoning_effort": "max",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/drizzle-orm-window-function-builders",
              "source_html_sha256": "d9285c39d82339465a6b8a6f99512f570092653f779e1f3e2ba8976c84acc346"
            }
          ],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "geo-shapeindex-serialization",
          "official_prompt": {
            "title": "Add ShapeIndex encoding and decoding",
            "description": "Add stable ShapeIndex Encode/Decode support so indices round-trip without rebuilding.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/geo-shapeindex-serialization",
            "instruction_sha256": "963f10993838f2f3f33935632d2b1181a1fa0b1174991a8b52ca8a66d40e5afd",
            "task_toml_sha256": "a95140adb066b732b69b8485392576cf89e0cf22a0d7028ad5d9d08f37c08f29",
            "language": "go",
            "category": "feature_request"
          },
          "repository": "https://github.com/golang/geo",
          "base_revision": "87f5a40ea07a4ea629ee5623c72660f3d1b217fa",
          "runs": [
            {
              "run_id": "tonight-20261012/geo-shapeindex-serialization__K72pa49",
              "task_id": "geo-shapeindex-serialization",
              "campaign": "tonight-20261012",
              "outcome": "pass",
              "score": {
                "feature": {
                  "passed": 24,
                  "total": 24
                },
                "preservation": {
                  "passed": 599,
                  "total": 599
                },
                "reward": 1
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "agent": "OpenCode/Pier",
                "runtime": "Strata 0.1.42",
                "agent_timeout_seconds": 2700
              },
              "timing_seconds": {
                "agent_execution": 604.371213,
                "environment_setup": 27.841622,
                "verifier": 30.706297
              },
              "usage": {
                "input_tokens": 2329387,
                "cached_tokens": 2269757,
                "output_tokens": 30405,
                "peak_context_tokens": 81336,
                "steps": 42,
                "requests": 43
              },
              "patch": {
                "bytes": 23848,
                "sha256": "06099ae4234e57469d58d2096202af1bf5fe583434ea23e4aebeaa5ae9472c18"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [
            {
              "model": "gpt-6-astra",
              "provider": "openai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_gpt_6_astra_xhigh",
              "reasoning_effort": "xhigh",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 2,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 3,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 4,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/geo-shapeindex-serialization",
              "source_html_sha256": "0325beec343d81448cae92a87beeef0a148d4d52424c59d6d083b3f713f89b42"
            },
            {
              "model": "claude-opus-5",
              "provider": "vertex_ai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_claude_opus_5_max",
              "reasoning_effort": "max",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/geo-shapeindex-serialization",
              "source_html_sha256": "0325beec343d81448cae92a87beeef0a148d4d52424c59d6d083b3f713f89b42"
            }
          ],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "wasmi-trap-coredumps",
          "official_prompt": {
            "title": "Add trap coredump generation to wasmi",
            "description": "Generate opt-in Wasm coredumps on traps and attach the bytes to errors.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/wasmi-trap-coredumps",
            "instruction_sha256": "119346d012f769e6a69d46ce176610fbf42a181122c2e85379d17abe56672888",
            "task_toml_sha256": "1e18db2a466b62290bb4fbf1b6653bed8302b7ae345a8db8c1b7d9770092c7ce",
            "language": "rust",
            "category": "feature_request"
          },
          "repository": "https://github.com/wasmi-labs/wasmi",
          "base_revision": "e1f76e285b9ad68a952b7cf5297bbb7ab91e6028",
          "runs": [
            {
              "run_id": "tonight-20261012/wasmi-trap-coredumps__SkX9dMo",
              "task_id": "wasmi-trap-coredumps",
              "campaign": "tonight-20261012",
              "outcome": "pass",
              "score": {
                "feature": {
                  "passed": 22,
                  "total": 22
                },
                "preservation": {
                  "passed": 58,
                  "total": 58
                },
                "reward": 1
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "agent": "OpenCode/Pier",
                "runtime": "Strata 0.1.42",
                "agent_timeout_seconds": 2700
              },
              "timing_seconds": {
                "agent_execution": 2477.977253,
                "environment_setup": 36.895853,
                "verifier": 26.534545
              },
              "usage": {
                "input_tokens": 15168388,
                "cached_tokens": 14921447,
                "output_tokens": 119100,
                "peak_context_tokens": 151972,
                "steps": 170,
                "requests": 171
              },
              "patch": {
                "bytes": 66852,
                "sha256": "d253f11edf1bcfe47ce6f332953fe83a57db9863004a9b2e777f62462a20dbac"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [
            {
              "model": "gpt-6-astra",
              "provider": "openai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_gpt_6_astra_xhigh",
              "reasoning_effort": "xhigh",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 2,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 3,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 4,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/wasmi-trap-coredumps",
              "source_html_sha256": "2678791ea7189e5315f5dbe054bcccb7bc57365b8715250cc6bbb0f2760f3454"
            },
            {
              "model": "claude-opus-5",
              "provider": "vertex_ai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_claude_opus_5_max",
              "reasoning_effort": "max",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/wasmi-trap-coredumps",
              "source_html_sha256": "2678791ea7189e5315f5dbe054bcccb7bc57365b8715250cc6bbb0f2760f3454"
            }
          ],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "yjs-map-conflict-detection",
          "official_prompt": {
            "title": "Add deterministic map conflict detection to Y.Map writes",
            "description": "Add strict, deterministic conflict detection for Y.Map key writes with collect and error policies.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/yjs-map-conflict-detection",
            "instruction_sha256": "ec6f54e53154aca614ead29fe45f2a85f876c4fe38aec5f3a0f20425e2611935",
            "task_toml_sha256": "efc0167671f665a71cbc94a9951fc1e85402e3a1efd20c285449606766a65428",
            "language": "javascript",
            "category": "feature_request"
          },
          "repository": "https://github.com/yjs/yjs",
          "base_revision": "7795050a749bd1111cbbdd9d0219b27226a8e710",
          "runs": [
            {
              "run_id": "tonight-20261012/yjs-map-conflict-detection__9oG3SR2",
              "task_id": "yjs-map-conflict-detection",
              "campaign": "tonight-20261012",
              "outcome": "pass",
              "score": {
                "feature": {
                  "passed": 9,
                  "total": 9
                },
                "preservation": {
                  "passed": 231,
                  "total": 231
                },
                "reward": 1
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "agent": "OpenCode/Pier",
                "runtime": "Strata 0.1.42",
                "agent_timeout_seconds": 2700
              },
              "timing_seconds": {
                "agent_execution": 882.088848,
                "environment_setup": 33.403535,
                "verifier": 25.034519
              },
              "usage": {
                "input_tokens": 5708118,
                "cached_tokens": 5636479,
                "output_tokens": 41470,
                "peak_context_tokens": 112705,
                "steps": 69,
                "requests": 70
              },
              "patch": {
                "bytes": 37459,
                "sha256": "43a486b3ee766060293274348b88b7539da633272e7d0a204627da62d35529ec"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [
            {
              "model": "gpt-6-astra",
              "provider": "openai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_gpt_6_astra_xhigh",
              "reasoning_effort": "xhigh",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 2,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 3,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 4,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/yjs-map-conflict-detection",
              "source_html_sha256": "cfd4cfce11cb4e61c85c683abbc1d260a127ff765b0df0b2f26acedae6c81edd"
            },
            {
              "model": "claude-opus-5",
              "provider": "vertex_ai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_claude_opus_5_max",
              "reasoning_effort": "max",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/yjs-map-conflict-detection",
              "source_html_sha256": "cfd4cfce11cb4e61c85c683abbc1d260a127ff765b0df0b2f26acedae6c81edd"
            }
          ],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "skrub-duration-encoding",
          "official_prompt": {
            "title": "Add duration encoding to TableVectorizer",
            "description": "Add a DurationEncoder, duration selector, and duration routing in TableVectorizer.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/skrub-duration-encoding",
            "instruction_sha256": "aecc575c2253564382af8fcb9afb0af11b47b8fda94c7625100a18148919e95d",
            "task_toml_sha256": "34e4539b4a1dad2374e5b4f1292295224b7e1002ce0754f5d7b47c69d8e1dc2e",
            "language": "python",
            "category": "feature_request"
          },
          "repository": "https://github.com/skrub-data/skrub",
          "base_revision": "24c4466fea94f551fb73d21eba54038dc5b346d3",
          "runs": [
            {
              "run_id": "tonight-20261012/skrub-duration-encoding__xiVc3Cb",
              "task_id": "skrub-duration-encoding",
              "campaign": "tonight-20261012",
              "outcome": "pass",
              "score": {
                "feature": {
                  "passed": 130,
                  "total": 130
                },
                "preservation": {
                  "passed": 2784,
                  "total": 2784
                },
                "reward": 1
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "agent": "OpenCode/Pier",
                "runtime": "Strata 0.1.42",
                "agent_timeout_seconds": 2700
              },
              "timing_seconds": {
                "agent_execution": 1475.190797,
                "environment_setup": 47.952636,
                "verifier": 279.391762
              },
              "usage": {
                "input_tokens": 9942671,
                "cached_tokens": 9863074,
                "output_tokens": 50986,
                "peak_context_tokens": 129831,
                "steps": 100,
                "requests": 101
              },
              "patch": {
                "bytes": 44206,
                "sha256": "f09f297e0dbd8ade7a207f1d8a73ef6801235fe216bc8360b22753083ae68997"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [
            {
              "model": "gpt-6-astra",
              "provider": "openai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_gpt_6_astra_xhigh",
              "reasoning_effort": "xhigh",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 2,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 3,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 4,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/skrub-duration-encoding",
              "source_html_sha256": "fa0696cb13c487e12d89ab60b7a288a0e37a6babe2ecef64d0bddf591eb7444f"
            },
            {
              "model": "claude-opus-5",
              "provider": "vertex_ai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_claude_opus_5_max",
              "reasoning_effort": "max",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/skrub-duration-encoding",
              "source_html_sha256": "fa0696cb13c487e12d89ab60b7a288a0e37a6babe2ecef64d0bddf591eb7444f"
            }
          ],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        }
      ]
    },
    {
      "id": "deepswe-followon-20261012",
      "label": "DeepSWE \u00b7 follow-on \u00b7 2026-10-12",
      "adapter": "deepswe",
      "benchmark": "DeepSWE v1.1",
      "campaign": "tonight-20261012-followon",
      "native_score_semantics": "feature-to-patch and patch-to-patch verifier checks",
      "tasks": [
        {
          "task_id": "abs-module-cache-flags",
          "official_prompt": {
            "title": "Harden module loading, cache introspection, and script flags",
            "description": "Harden ABS module resolution and caching, expose cache introspection APIs, and make module flags work correctly in script mode.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/abs-module-cache-flags",
            "instruction_sha256": "cffcaa70386408363b090b563943fd63fdcc322ed44b8c76cc3e56cfd76b7ed8",
            "task_toml_sha256": "bb44f3275fde199bbe5d6706630a5c6140cfe5beb66e49f2c391fa548bd974e6",
            "language": "go",
            "category": "enhancement"
          },
          "repository": "https://github.com/abs-lang/abs",
          "base_revision": "cb1b3b671d0ee9fa9da9f7b02f86967953ffd10a",
          "runs": [
            {
              "run_id": "tonight-20261012-followon/abs-module-cache-flags__eLaT3ZE",
              "task_id": "abs-module-cache-flags",
              "campaign": "tonight-20261012-followon",
              "outcome": "graded-model-failure",
              "score": {
                "feature": {
                  "passed": 13,
                  "total": 20
                },
                "preservation": {
                  "passed": 3,
                  "total": 3
                },
                "reward": 0
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "agent": "OpenCode/Pier",
                "runtime": "Strata 0.1.42",
                "agent_timeout_seconds": 2700
              },
              "timing_seconds": {
                "agent_execution": 511.82121,
                "environment_setup": 3.665557,
                "verifier": 30.864019
              },
              "usage": {
                "input_tokens": 1892538,
                "cached_tokens": 1868392,
                "output_tokens": 22798,
                "peak_context_tokens": 46598,
                "steps": 63,
                "requests": 64
              },
              "patch": {
                "bytes": 19487,
                "sha256": "b0e8090ad4178406cfad69a6c73fd01c4382c87c26f7301f96ae399d547dc645"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [
            {
              "model": "gpt-6-astra",
              "provider": "openai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_gpt_6_astra_xhigh",
              "reasoning_effort": "xhigh",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 2,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 3,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 4,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/abs-module-cache-flags",
              "source_html_sha256": "ec83e8c22cd6a032d9bfd20aac9ad15dfba54ba98674abd5c52409b7ce61d671"
            },
            {
              "model": "claude-opus-5",
              "provider": "vertex_ai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_claude_opus_5_max",
              "reasoning_effort": "max",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/abs-module-cache-flags",
              "source_html_sha256": "ec83e8c22cd6a032d9bfd20aac9ad15dfba54ba98674abd5c52409b7ce61d671"
            }
          ],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "adaptix-name-mapping-aliases",
          "official_prompt": {
            "title": "Add input key aliases to name mapping",
            "description": "Add load-only alias support to name mapping so fields can resolve from alternate input keys.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/adaptix-name-mapping-aliases",
            "instruction_sha256": "e8e4a39e65723ff56eb35bc9b9183fef4e09677159063abf8189f471cc1b2eac",
            "task_toml_sha256": "ec03c8bc5fd8ba4d86188d4b92681c4454bb90376ac4af1fdfda6353f9201b3f",
            "language": "python",
            "category": "feature_request"
          },
          "repository": "https://github.com/reagento/adaptix",
          "base_revision": "a691069fcadf9131e5f7a5a130a022dc678f3e1d",
          "runs": [
            {
              "run_id": "tonight-20261012-followon/adaptix-name-mapping-aliases__zgYeMuM",
              "task_id": "adaptix-name-mapping-aliases",
              "campaign": "tonight-20261012-followon",
              "outcome": "pass",
              "score": {
                "feature": {
                  "passed": 44,
                  "total": 44
                },
                "preservation": {
                  "passed": 2738,
                  "total": 2738
                },
                "reward": 1
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "agent": "OpenCode/Pier",
                "runtime": "Strata 0.1.42",
                "agent_timeout_seconds": 2700
              },
              "timing_seconds": {
                "agent_execution": 1639.263037,
                "environment_setup": 26.016893,
                "verifier": 32.357342
              },
              "usage": {
                "input_tokens": 12205840,
                "cached_tokens": 12126886,
                "output_tokens": 79318,
                "peak_context_tokens": 155854,
                "steps": 128,
                "requests": 129
              },
              "patch": {
                "bytes": 36614,
                "sha256": "3102729c954eb89ba07e4969fc906de99ebe55684df2f2d39e9625b77fbf7e44"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [
            {
              "model": "gpt-6-astra",
              "provider": "openai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_gpt_6_astra_xhigh",
              "reasoning_effort": "xhigh",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 2,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 3,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 4,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/adaptix-name-mapping-aliases",
              "source_html_sha256": "f765ac2f9afed1a8cd683b5ee370d82e5ea16c1559f4e8ca2c5a09c4e6c08106"
            },
            {
              "model": "claude-opus-5",
              "provider": "vertex_ai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_claude_opus_5_max",
              "reasoning_effort": "max",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/adaptix-name-mapping-aliases",
              "source_html_sha256": "f765ac2f9afed1a8cd683b5ee370d82e5ea16c1559f4e8ca2c5a09c4e6c08106"
            }
          ],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "bandit-incremental-cache-control",
          "official_prompt": {
            "title": "Add incremental cache controls to Bandit",
            "description": "Add incremental analysis caching with cache invalidation, import/export, pruning, and cache inspection CLI options.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/bandit-incremental-cache-control",
            "instruction_sha256": "78a166bd4aca7057e98d11106812396f9aa695a80b57ce346aa8231af2cecb12",
            "task_toml_sha256": "1807632d919182dd98c7ba8398ef82f8d0982b12739c0b209322130893424a75",
            "language": "python",
            "category": "feature_request"
          },
          "repository": "https://github.com/PyCQA/bandit.git",
          "base_revision": "765f00d3f202f83f61d03f882f80a2d5142d81f8",
          "runs": [
            {
              "run_id": "tonight-20261012-followon/bandit-incremental-cache-control__RyHWf2U",
              "task_id": "bandit-incremental-cache-control",
              "campaign": "tonight-20261012-followon",
              "outcome": "pass",
              "score": {
                "feature": {
                  "passed": 88,
                  "total": 88
                },
                "preservation": {
                  "passed": 275,
                  "total": 275
                },
                "reward": 1
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "agent": "OpenCode/Pier",
                "runtime": "Strata 0.1.42",
                "agent_timeout_seconds": 2700
              },
              "timing_seconds": {
                "agent_execution": 708.818779,
                "environment_setup": 25.289135,
                "verifier": 49.41769
              },
              "usage": {
                "input_tokens": 3472125,
                "cached_tokens": 3430746,
                "output_tokens": 36265,
                "peak_context_tokens": 77258,
                "steps": 62,
                "requests": 63
              },
              "patch": {
                "bytes": 50359,
                "sha256": "53f19a837f4d76a5b2dc973a130e78e6a9201e808510df2ebff3f53c593d8f18"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [
            {
              "model": "gpt-6-astra",
              "provider": "openai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_gpt_6_astra_xhigh",
              "reasoning_effort": "xhigh",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 2,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 3,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 4,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/bandit-incremental-cache-control",
              "source_html_sha256": "6ae477b81c5ea43969229e31a08e38d70448ea80509c9fbaecbc994341082de9"
            },
            {
              "model": "claude-opus-5",
              "provider": "vertex_ai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_claude_opus_5_max",
              "reasoning_effort": "max",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/bandit-incremental-cache-control",
              "source_html_sha256": "6ae477b81c5ea43969229e31a08e38d70448ea80509c9fbaecbc994341082de9"
            }
          ],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "boa-hierarchical-evaluation-cancellation",
          "official_prompt": {
            "title": "Add hierarchical evaluation cancellation to Boa",
            "description": "Add cancellable evaluation handles that propagate through nested script, module, and job execution.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/boa-hierarchical-evaluation-cancellation",
            "instruction_sha256": "72e6aa910687e1fe0401b7d97183880ee143d60b5183cb554d554abe003cedec",
            "task_toml_sha256": "01f6340fa9c876a0707f3a71a1196820b2de88d0ae292b7778d5acd3407d1dab",
            "language": "rust",
            "category": "feature_request"
          },
          "repository": "https://github.com/boa-dev/boa",
          "base_revision": "70409a5052984325dccfdc5f6520818568a81f39",
          "runs": [
            {
              "run_id": "tonight-20261012-followon/boa-hierarchical-evaluation-canc__yCbepPx",
              "task_id": "boa-hierarchical-evaluation-cancellation",
              "campaign": "tonight-20261012-followon",
              "outcome": "graded-model-failure",
              "score": {
                "feature": {
                  "passed": 0,
                  "total": 17
                },
                "preservation": {
                  "passed": 7,
                  "total": 7
                },
                "reward": 0
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "agent": "OpenCode/Pier",
                "runtime": "Strata 0.1.42",
                "agent_timeout_seconds": 2700
              },
              "timing_seconds": {
                "agent_execution": 2009.5824,
                "environment_setup": 64.02786,
                "verifier": 80.237243
              },
              "usage": {
                "input_tokens": 10469212,
                "cached_tokens": 10265305,
                "output_tokens": 95675,
                "peak_context_tokens": 156825,
                "steps": 128,
                "requests": 129
              },
              "patch": {
                "bytes": 54560,
                "sha256": "7ffe34ccd85940b3c5e2d08feea5d0d6accb83d9d865308ddf56629b39924ad2"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [
            {
              "model": "gpt-6-astra",
              "provider": "openai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_gpt_6_astra_xhigh",
              "reasoning_effort": "xhigh",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 2,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 3,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 4,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/boa-hierarchical-evaluation-cancellation",
              "source_html_sha256": "c0376962644950b0f75be082a958f7a7ab888f64bfdc1e0a3b068b0edebdf0b1"
            },
            {
              "model": "claude-opus-5",
              "provider": "vertex_ai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_claude_opus_5_max",
              "reasoning_effort": "max",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/boa-hierarchical-evaluation-cancellation",
              "source_html_sha256": "c0376962644950b0f75be082a958f7a7ab888f64bfdc1e0a3b068b0edebdf0b1"
            }
          ],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "cliffy-config-file-parsing",
          "official_prompt": {
            "title": "Add config file parsing to Cliffy commands",
            "description": "Add command-level config file loading, parsing, merging, and precedence handling.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/cliffy-config-file-parsing",
            "instruction_sha256": "39668d18670621b7760751128fc57ffbe519e3529601567ea9e11d31faaa2daf",
            "task_toml_sha256": "c21651d7a9fd6b3dbe5562e804153c51775cf1d4ed112f775229d11560b94ba4",
            "language": "typescript",
            "category": "feature_request"
          },
          "repository": "https://github.com/c4spar/cliffy",
          "base_revision": "132a437c40cffbdfbe474ca808c8debde59e2633",
          "runs": [
            {
              "run_id": "tonight-20261012-followon/cliffy-config-file-parsing__pGRckyX",
              "task_id": "cliffy-config-file-parsing",
              "campaign": "tonight-20261012-followon",
              "outcome": "pass",
              "score": {
                "feature": {
                  "passed": 37,
                  "total": 37
                },
                "preservation": {
                  "passed": 451,
                  "total": 451
                },
                "reward": 1
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "agent": "OpenCode/Pier",
                "runtime": "Strata 0.1.42",
                "agent_timeout_seconds": 2700
              },
              "timing_seconds": {
                "agent_execution": 751.083125,
                "environment_setup": 26.391629,
                "verifier": 22.103584
              },
              "usage": {
                "input_tokens": 4385643,
                "cached_tokens": 4319903,
                "output_tokens": 38157,
                "peak_context_tokens": 103588,
                "steps": 57,
                "requests": 58
              },
              "patch": {
                "bytes": 33738,
                "sha256": "0734e220d0285e4fbcf72b9c7a31769bfa73c1049a14f27c4a94ffcecf978d20"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [
            {
              "model": "gpt-6-astra",
              "provider": "openai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_gpt_6_astra_xhigh",
              "reasoning_effort": "xhigh",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 2,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 3,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 4,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/cliffy-config-file-parsing",
              "source_html_sha256": "13811f25a14e903c07de301b5a598ca586fbcec89a1cda819e9769c9da1fedec"
            },
            {
              "model": "claude-opus-5",
              "provider": "vertex_ai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_claude_opus_5_max",
              "reasoning_effort": "max",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 3,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/cliffy-config-file-parsing",
              "source_html_sha256": "13811f25a14e903c07de301b5a598ca586fbcec89a1cda819e9769c9da1fedec"
            }
          ],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "dasel-html-document-format",
          "official_prompt": {
            "title": "Add HTML document format handling to Dasel",
            "description": "Add read and write support for HTML documents with Dasel's format handling.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/dasel-html-document-format",
            "instruction_sha256": "53a67fd1336d4b281f47b37e809834369ebd9d1d2178c5da2c80fd1b0e4393f8",
            "task_toml_sha256": "e178607fca1ac99213f24493abcdbc97c1258decbbd16ee660adc35f29538443",
            "language": "go",
            "category": "feature_request"
          },
          "repository": "https://github.com/TomWright/dasel",
          "base_revision": "0dd6132e0c58edbd9b1a5f7ffd00dfab1e6085ad",
          "runs": [
            {
              "run_id": "tonight-20261012-followon/dasel-html-document-format__zHnrBQG",
              "task_id": "dasel-html-document-format",
              "campaign": "tonight-20261012-followon",
              "outcome": "graded-model-failure",
              "score": {
                "feature": {
                  "passed": 144,
                  "total": 146
                },
                "preservation": {
                  "passed": 1012,
                  "total": 1012
                },
                "reward": 0
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "agent": "OpenCode/Pier",
                "runtime": "Strata 0.1.42",
                "agent_timeout_seconds": 2700
              },
              "timing_seconds": {
                "agent_execution": 968.449202,
                "environment_setup": 25.646378,
                "verifier": 36.850513
              },
              "usage": {
                "input_tokens": 3855429,
                "cached_tokens": 3816488,
                "output_tokens": 50558,
                "peak_context_tokens": 83376,
                "steps": 69,
                "requests": 70
              },
              "patch": {
                "bytes": 45138,
                "sha256": "054498882e7044f3070732a74f084be1001c0b7fd37479eedf786d2d14fab63c"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [
            {
              "model": "gpt-6-astra",
              "provider": "openai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_gpt_6_astra_xhigh",
              "reasoning_effort": "xhigh",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 2,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 3,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 4,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/dasel-html-document-format",
              "source_html_sha256": "7826855f3723b386ae723621235eb393698d31490b77a6934506f34530d98386"
            },
            {
              "model": "claude-opus-5",
              "provider": "vertex_ai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_claude_opus_5_max",
              "reasoning_effort": "max",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/dasel-html-document-format",
              "source_html_sha256": "7826855f3723b386ae723621235eb393698d31490b77a6934506f34530d98386"
            }
          ],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "dynamodb-toolbox-conditional-attribute-requirements",
          "official_prompt": {
            "title": "Add conditional required attributes to schemas",
            "description": "Add `requiredIf`-based conditional attribute enforcement across schema validation, parsing, updates, and JSON Schema export.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/dynamodb-toolbox-conditional-attribute-requirements",
            "instruction_sha256": "b3f779ea7ac151568f0dad3838f9bcdeabff7e4fca45677b26c406fa442a7fc6",
            "task_toml_sha256": "06726026383a8949dba3e40e939a6064754cb0c94068985ab589acd0021a4ee5",
            "language": "typescript",
            "category": "feature_request"
          },
          "repository": "https://github.com/dynamodb-toolbox/dynamodb-toolbox",
          "base_revision": "1f2a18664f8aded292707fcafb01ff15ea33d3b8",
          "runs": [
            {
              "run_id": "tonight-20261012-followon/dynamodb-toolbox-conditional-att__iuC7yiG",
              "task_id": "dynamodb-toolbox-conditional-attribute-requirements",
              "campaign": "tonight-20261012-followon",
              "outcome": "graded-model-failure",
              "score": {
                "feature": {
                  "passed": 29,
                  "total": 31
                },
                "preservation": {
                  "passed": 1267,
                  "total": 1267
                },
                "reward": 0
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "agent": "OpenCode/Pier",
                "runtime": "Strata 0.1.42",
                "agent_timeout_seconds": 2700
              },
              "timing_seconds": {
                "agent_execution": 2001.096069,
                "environment_setup": 29.343243,
                "verifier": 42.650691
              },
              "usage": {
                "input_tokens": 9721731,
                "cached_tokens": 9635524,
                "output_tokens": 64159,
                "peak_context_tokens": 140224,
                "steps": 100,
                "requests": 155
              },
              "patch": {
                "bytes": 70150,
                "sha256": "9a36dbfa14938a6258438c4d702af063afadc8910e07a1ccd411ba4664b6beb8"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [
            {
              "model": "gpt-6-astra",
              "provider": "openai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_gpt_6_astra_xhigh",
              "reasoning_effort": "xhigh",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 2,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 3,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 4,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/dynamodb-toolbox-conditional-attribute-requirements",
              "source_html_sha256": "b5164e24d058b7a709b2c298578896f38760af2a4bd79cf1b0b126151f55e32a"
            },
            {
              "model": "claude-opus-5",
              "provider": "vertex_ai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_claude_opus_5_max",
              "reasoning_effort": "max",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 2,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/dynamodb-toolbox-conditional-attribute-requirements",
              "source_html_sha256": "b5164e24d058b7a709b2c298578896f38760af2a4bd79cf1b0b126151f55e32a"
            }
          ],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "fastapi-implicit-head-options",
          "official_prompt": {
            "title": "Add implicit HEAD and automatic OPTIONS responses to FastAPI routes",
            "description": "Add configurable implicit HEAD handling and automatic OPTIONS responses for FastAPI routes, routers, and included routers.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/fastapi-implicit-head-options",
            "instruction_sha256": "b55a492424e74cf6267a1bcf75c48942f397e9e9f00ac15e97809145794056d6",
            "task_toml_sha256": "4d48ce45311547285d7ad281ecfd5278abf73d3355305bce88bb889e99980f89",
            "language": "python",
            "category": "feature_request"
          },
          "repository": "https://github.com/fastapi/fastapi",
          "base_revision": "11614be9021aa4ac078d4d0693a8b5250a1010d8",
          "runs": [
            {
              "run_id": "tonight-20261012-followon/fastapi-implicit-head-options__bpZacWE",
              "task_id": "fastapi-implicit-head-options",
              "campaign": "tonight-20261012-followon",
              "outcome": "pass",
              "score": {
                "feature": {
                  "passed": 43,
                  "total": 43
                },
                "preservation": {
                  "passed": 3134,
                  "total": 3134
                },
                "reward": 1
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "agent": "OpenCode/Pier",
                "runtime": "Strata 0.1.42",
                "agent_timeout_seconds": 2700
              },
              "timing_seconds": {
                "agent_execution": 1385.555989,
                "environment_setup": 23.871943,
                "verifier": 60.917291
              },
              "usage": {
                "input_tokens": 8925310,
                "cached_tokens": 8862057,
                "output_tokens": 66984,
                "peak_context_tokens": 124693,
                "steps": 111,
                "requests": 112
              },
              "patch": {
                "bytes": 76868,
                "sha256": "609ce41c581e48b2a002632fcf9394de0e3c2dd33c638ea87d9564f8a7a96af3"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [
            {
              "model": "gpt-6-astra",
              "provider": "openai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_gpt_6_astra_xhigh",
              "reasoning_effort": "xhigh",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 2,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 3,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 4,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/fastapi-implicit-head-options",
              "source_html_sha256": "42633a6363d0d5de626a90e11329ee9dab5bb3ecb331a27600d585dc1bd2561a"
            },
            {
              "model": "claude-opus-5",
              "provider": "vertex_ai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_claude_opus_5_max",
              "reasoning_effort": "max",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/fastapi-implicit-head-options",
              "source_html_sha256": "42633a6363d0d5de626a90e11329ee9dab5bb3ecb331a27600d585dc1bd2561a"
            }
          ],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "go-genai-streamed-function-args",
          "official_prompt": {
            "title": "Expose accumulated streamed function-call args in SDK surfaces",
            "description": "Expose fully accumulated streamed function-call arguments across streaming, live sessions, and chat history.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/go-genai-streamed-function-args",
            "instruction_sha256": "a01fe3132affb9ec483885da4fd8574d3dd056cf7448a2a7cdc74e6486aed632",
            "task_toml_sha256": "0a5f6124b68f5938ae0315079a7a2e7d0aa62ece38e4ad380a4a18703696b537",
            "language": "go",
            "category": "feature_request"
          },
          "repository": "https://github.com/googleapis/go-genai.git",
          "base_revision": "87c0e5a4f27d04569d927717769f34483e0ba475",
          "runs": [
            {
              "run_id": "tonight-20261012-followon/go-genai-streamed-function-args__9GjSjEG",
              "task_id": "go-genai-streamed-function-args",
              "campaign": "tonight-20261012-followon",
              "outcome": "pass",
              "score": {
                "feature": {
                  "passed": 6,
                  "total": 6
                },
                "preservation": {
                  "passed": 62,
                  "total": 62
                },
                "reward": 1
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "agent": "OpenCode/Pier",
                "runtime": "Strata 0.1.42",
                "agent_timeout_seconds": 2700
              },
              "timing_seconds": {
                "agent_execution": 1022.424934,
                "environment_setup": 24.633935,
                "verifier": 38.194232
              },
              "usage": {
                "input_tokens": 5318826,
                "cached_tokens": 5262030,
                "output_tokens": 50785,
                "peak_context_tokens": 97811,
                "steps": 92,
                "requests": 93
              },
              "patch": {
                "bytes": 41684,
                "sha256": "7d72e94aaa631fadb38b0c270a2e3449f6dd38b0e0f131c99e09157edfbe53ff"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [
            {
              "model": "gpt-6-astra",
              "provider": "openai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_gpt_6_astra_xhigh",
              "reasoning_effort": "xhigh",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 2,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 3,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 4,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/go-genai-streamed-function-args",
              "source_html_sha256": "99e17cd6387a7e3fed46915adc5b09dc8552dff861b2b395ee693db72424e1d9"
            },
            {
              "model": "claude-opus-5",
              "provider": "vertex_ai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_claude_opus_5_max",
              "reasoning_effort": "max",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/go-genai-streamed-function-args",
              "source_html_sha256": "99e17cd6387a7e3fed46915adc5b09dc8552dff861b2b395ee693db72424e1d9"
            }
          ],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "go-git-worktree-merge-conflicts",
          "official_prompt": {
            "title": "Add worktree merge conflict handling",
            "description": "Add worktree merge support with conflict detection, merge head handling, and index stage updates.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/go-git-worktree-merge-conflicts",
            "instruction_sha256": "4d6c197ad66c5d90b51f52d7db5f539fd164ce9b60e6c0d8bfa356f1f77cea7e",
            "task_toml_sha256": "17a105a1d8e8aed5f45da06820dc65a1048a3a3f5a8d0f2dbd7c23ee74f94eb9",
            "language": "go",
            "category": "feature_request"
          },
          "repository": "https://github.com/go-git/go-git",
          "base_revision": "424e9964d3a33c6507a77c126841f2c5897262af",
          "runs": [
            {
              "run_id": "tonight-20261012-followon/go-git-worktree-merge-conflicts__6DfcCJY",
              "task_id": "go-git-worktree-merge-conflicts",
              "campaign": "tonight-20261012-followon",
              "outcome": "pass",
              "score": {
                "feature": {
                  "passed": 17,
                  "total": 17
                },
                "preservation": {
                  "passed": 2,
                  "total": 2
                },
                "reward": 1
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "agent": "OpenCode/Pier",
                "runtime": "Strata 0.1.42",
                "agent_timeout_seconds": 2700
              },
              "timing_seconds": {
                "agent_execution": 958.982545,
                "environment_setup": 29.301627,
                "verifier": 38.838269
              },
              "usage": {
                "input_tokens": 3175660,
                "cached_tokens": 3111087,
                "output_tokens": 49403,
                "peak_context_tokens": 101860,
                "steps": 48,
                "requests": 49
              },
              "patch": {
                "bytes": 39034,
                "sha256": "8ba982f0d1cdb57134a0e6d9df9a6d1a0a31e4b01b7cae6466a41817da47516c"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [
            {
              "model": "gpt-6-astra",
              "provider": "openai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_gpt_6_astra_xhigh",
              "reasoning_effort": "xhigh",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 2,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 3,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 4,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/go-git-worktree-merge-conflicts",
              "source_html_sha256": "7bce4050373ccf787578ca6686727c9b798db2f9887e1ad8c2c2ddeba8b0c044"
            },
            {
              "model": "claude-opus-5",
              "provider": "vertex_ai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_claude_opus_5_max",
              "reasoning_effort": "max",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/go-git-worktree-merge-conflicts",
              "source_html_sha256": "7bce4050373ccf787578ca6686727c9b798db2f9887e1ad8c2c2ddeba8b0c044"
            }
          ],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "goreleaser-retry-publish-auditing",
          "official_prompt": {
            "title": "Add retry-aware publishing audit logs",
            "description": "Implement per-artifact retries for uploads, artifactory, and blobs while recording deterministic publish attempt history.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/goreleaser-retry-publish-auditing",
            "instruction_sha256": "b6474aa4d752c52c130bb2d4627ed5742caabb685570f689bc756fd8b8baa128",
            "task_toml_sha256": "deddcda00d669f9962eec4cdd418549a54c6ad2e51cf52cbcd253637e5f1d9c1",
            "language": "go",
            "category": "feature_request"
          },
          "repository": "https://github.com/goreleaser/goreleaser",
          "base_revision": "399ef141161f212f4e81b5d7497b84633fc712d9",
          "runs": [
            {
              "run_id": "tonight-20261012-followon/goreleaser-retry-publish-auditin__HiH3HZt",
              "task_id": "goreleaser-retry-publish-auditing",
              "campaign": "tonight-20261012-followon",
              "outcome": "graded-model-failure",
              "score": {
                "feature": {
                  "passed": 28,
                  "total": 29
                },
                "preservation": {
                  "passed": 29,
                  "total": 29
                },
                "reward": 0
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "agent": "OpenCode/Pier",
                "runtime": "Strata 0.1.42",
                "agent_timeout_seconds": 2700
              },
              "timing_seconds": {
                "agent_execution": 1296.807556,
                "environment_setup": 104.297679,
                "verifier": 70.727088
              },
              "usage": {
                "input_tokens": 5935433,
                "cached_tokens": 5894142,
                "output_tokens": 50185,
                "peak_context_tokens": 90616,
                "steps": 107,
                "requests": 108
              },
              "patch": {
                "bytes": 47538,
                "sha256": "0fb2757d1bddb8ac1fd3eece73cd731335407eb0608876643f89b6dfb44d2374"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [
            {
              "model": "gpt-6-astra",
              "provider": "openai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_gpt_6_astra_xhigh",
              "reasoning_effort": "xhigh",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 2,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 3,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 4,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/goreleaser-retry-publish-auditing",
              "source_html_sha256": "9fd4720af3728a0a911efab806b640c2501ec615783be106ffaae30d76ebfc30"
            },
            {
              "model": "claude-opus-5",
              "provider": "vertex_ai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_claude_opus_5_max",
              "reasoning_effort": "max",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/goreleaser-retry-publish-auditing",
              "source_html_sha256": "9fd4720af3728a0a911efab806b640c2501ec615783be106ffaae30d76ebfc30"
            }
          ],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "helm-unified-manifest-stream",
          "official_prompt": {
            "title": "Add unified manifest stream output across Helm commands",
            "description": "Add a stable unified manifest stream for template, dry-run install/upgrade, and get manifest output.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/helm-unified-manifest-stream",
            "instruction_sha256": "c69a568df4d740e1e4c6b1e8db40d9fcb3562fafc2f2c3422daf53181b808799",
            "task_toml_sha256": "2660916e53c102665b429f50173ee87dad21a319a4e6aea60900e24d4b8b3406",
            "language": "go",
            "category": "feature_request"
          },
          "repository": "https://github.com/helm/helm",
          "base_revision": "42f78ba60edf531d5161e00d9819a7c34d976343",
          "runs": [
            {
              "run_id": "tonight-20261012-followon/helm-unified-manifest-stream__K5gXqBY",
              "task_id": "helm-unified-manifest-stream",
              "campaign": "tonight-20261012-followon",
              "outcome": "graded-model-failure",
              "score": {
                "feature": {
                  "passed": 0,
                  "total": 5
                },
                "preservation": {
                  "passed": 2,
                  "total": 2
                },
                "reward": 0
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "agent": "OpenCode/Pier",
                "runtime": "Strata 0.1.42",
                "agent_timeout_seconds": 2700
              },
              "timing_seconds": {
                "agent_execution": 2203.952954,
                "environment_setup": 3.461494,
                "verifier": 103.146148
              },
              "usage": {
                "input_tokens": 13023281,
                "cached_tokens": 12793433,
                "output_tokens": 82139,
                "peak_context_tokens": 156838,
                "steps": 154,
                "requests": 155
              },
              "patch": {
                "bytes": 0,
                "sha256": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [
            {
              "model": "gpt-6-astra",
              "provider": "openai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_gpt_6_astra_xhigh",
              "reasoning_effort": "xhigh",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 2,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 3,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 4,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/helm-unified-manifest-stream",
              "source_html_sha256": "b9019b62500c47fd63e850f155603ead8e6c0d6249170f9d9bd596310ec25cd1"
            },
            {
              "model": "claude-opus-5",
              "provider": "vertex_ai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_claude_opus_5_max",
              "reasoning_effort": "max",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 2,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/helm-unified-manifest-stream",
              "source_html_sha256": "b9019b62500c47fd63e850f155603ead8e6c0d6249170f9d9bd596310ec25cd1"
            }
          ],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "httpx-deterministic-cookie-store",
          "official_prompt": {
            "title": "Add a deterministic CookieStore with modern Set-Cookie parsing",
            "description": "Add a deterministic CookieStore with modern Set-Cookie parsing, cookie eviction, and request header handling.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/httpx-deterministic-cookie-store",
            "instruction_sha256": "97bd72180711c083997403fd5e88d7ebcca405360b06b93d5080cd2a6e625f22",
            "task_toml_sha256": "625136dd58235803eb622abda1855fdc4dc790b4016ce51c9ab987a8067f0b5a",
            "language": "typescript",
            "category": "feature_request"
          },
          "repository": "https://github.com/encode/httpx",
          "base_revision": "b5addb64f0161ff6bfe94c124ef76f6a1fba5254",
          "runs": [
            {
              "run_id": "tonight-20261012-followon/httpx-deterministic-cookie-store__9F5go7T",
              "task_id": "httpx-deterministic-cookie-store",
              "campaign": "tonight-20261012-followon",
              "outcome": "graded-model-failure",
              "score": {
                "feature": {
                  "passed": 114,
                  "total": 115
                },
                "preservation": {
                  "passed": 1281,
                  "total": 1281
                },
                "reward": 0
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "agent": "OpenCode/Pier",
                "runtime": "Strata 0.1.42",
                "agent_timeout_seconds": 2700
              },
              "timing_seconds": {
                "agent_execution": 722.03391,
                "environment_setup": 3.742121,
                "verifier": 22.777874
              },
              "usage": {
                "input_tokens": 2198191,
                "cached_tokens": 2176076,
                "output_tokens": 32084,
                "peak_context_tokens": 53651,
                "steps": 58,
                "requests": 59
              },
              "patch": {
                "bytes": 40321,
                "sha256": "6daed644994b555a569adbd025079a8a8df8210e6c164c084a1ba198793bf83b"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [
            {
              "model": "gpt-6-astra",
              "provider": "openai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_gpt_6_astra_xhigh",
              "reasoning_effort": "xhigh",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 2,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 3,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 4,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/httpx-deterministic-cookie-store",
              "source_html_sha256": "a489229654746a123330465bfcf8285e4b6600de78acbf42f6e01b8acc78be1d"
            },
            {
              "model": "claude-opus-5",
              "provider": "vertex_ai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_claude_opus_5_max",
              "reasoning_effort": "max",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/httpx-deterministic-cookie-store",
              "source_html_sha256": "a489229654746a123330465bfcf8285e4b6600de78acbf42f6e01b8acc78be1d"
            }
          ],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "httpx-multipart-response-parsing",
          "official_prompt": {
            "title": "Add multipart response parsing to HTTPX",
            "description": "Add Response iterators that parse multipart HTTP response bodies into parts.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/httpx-multipart-response-parsing",
            "instruction_sha256": "6a9c3400a23ad61f43c6e6f5b2dfbe8f5a5af81061cc063fd9440f327c5a0fb5",
            "task_toml_sha256": "7ee0209ec928035b9c0ac28997c886cbef8197567429e1c0493631856fdad5e5",
            "language": "python",
            "category": "feature_request"
          },
          "repository": "https://github.com/encode/httpx",
          "base_revision": "b5addb64f0161ff6bfe94c124ef76f6a1fba5254",
          "runs": [
            {
              "run_id": "tonight-20261012-followon/httpx-multipart-response-parsing__7K8NvzG",
              "task_id": "httpx-multipart-response-parsing",
              "campaign": "tonight-20261012-followon",
              "outcome": "pass",
              "score": {
                "feature": {
                  "passed": 122,
                  "total": 122
                },
                "preservation": {
                  "passed": 1272,
                  "total": 1272
                },
                "reward": 1
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "agent": "OpenCode/Pier",
                "runtime": "Strata 0.1.42",
                "agent_timeout_seconds": 2700
              },
              "timing_seconds": {
                "agent_execution": 705.176334,
                "environment_setup": 3.5341,
                "verifier": 22.755236
              },
              "usage": {
                "input_tokens": 2314026,
                "cached_tokens": 2286306,
                "output_tokens": 35259,
                "peak_context_tokens": 62440,
                "steps": 53,
                "requests": 54
              },
              "patch": {
                "bytes": 25954,
                "sha256": "1242d7fe549b4bd28bdcdb56d9bbb9962d814455f30b715f550239f6d9aaf379"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [
            {
              "model": "gpt-6-astra",
              "provider": "openai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_gpt_6_astra_xhigh",
              "reasoning_effort": "xhigh",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 2,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 3,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 4,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/httpx-multipart-response-parsing",
              "source_html_sha256": "c15caa82284a81eda38f522f3af877ba6eaa2c13c9aa255bba4b909c825ed39d"
            },
            {
              "model": "claude-opus-5",
              "provider": "vertex_ai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_claude_opus_5_max",
              "reasoning_effort": "max",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/httpx-multipart-response-parsing",
              "source_html_sha256": "c15caa82284a81eda38f522f3af877ba6eaa2c13c9aa255bba4b909c825ed39d"
            }
          ],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "ipython-session-bundle-replay",
          "official_prompt": {
            "title": "Add session bundle recording and replay to IPython",
            "description": "Add `.ipybundle` recording, validation, and replay APIs for IPython sessions.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/ipython-session-bundle-replay",
            "instruction_sha256": "4f32a1ba02f7c46ccdedd09b6de17a91a89b0567980bb343508ecdfcd90388b1",
            "task_toml_sha256": "eaf41b8bf1a7965f6d050b6b4c3799a919871461a29dab6fc681852eba9a5417",
            "language": "python",
            "category": "feature_request"
          },
          "repository": "https://github.com/ipython/ipython",
          "base_revision": "0bb317d10fdcb3aa13beb1031d5f10e5b821203b",
          "runs": [
            {
              "run_id": "tonight-20261012-followon/ipython-session-bundle-replay__3pWUJCQ",
              "task_id": "ipython-session-bundle-replay",
              "campaign": "tonight-20261012-followon",
              "outcome": "graded-model-failure",
              "score": {
                "feature": {
                  "passed": 14,
                  "total": 17
                },
                "preservation": {
                  "passed": 29,
                  "total": 29
                },
                "reward": 0
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "agent": "OpenCode/Pier",
                "runtime": "Strata 0.1.42",
                "agent_timeout_seconds": 2700
              },
              "timing_seconds": {
                "agent_execution": 692.884049,
                "environment_setup": 25.091993,
                "verifier": 18.393327
              },
              "usage": {
                "input_tokens": 1902257,
                "cached_tokens": 1878029,
                "output_tokens": 31843,
                "peak_context_tokens": 54809,
                "steps": 58,
                "requests": 59
              },
              "patch": {
                "bytes": 43116,
                "sha256": "7e827ffc5157c31e05deafafeccfa138306312d50255d27514b9f7dbb5bbe9fa"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [
            {
              "model": "gpt-6-astra",
              "provider": "openai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_gpt_6_astra_xhigh",
              "reasoning_effort": "xhigh",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 2,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 3,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 4,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/ipython-session-bundle-replay",
              "source_html_sha256": "2119ac8a39fae406f69148d8bf07aaa01c780d90b0002e427e6768a0c79e4f1c"
            },
            {
              "model": "claude-opus-5",
              "provider": "vertex_ai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_claude_opus_5_max",
              "reasoning_effort": "max",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/ipython-session-bundle-replay",
              "source_html_sha256": "2119ac8a39fae406f69148d8bf07aaa01c780d90b0002e427e6768a0c79e4f1c"
            }
          ],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "langchain-request-coalescing",
          "official_prompt": {
            "title": "Add request coalescing to `Runnable`",
            "description": "Add `Runnable.with_coalesce()` so concurrent identical inputs share one execution across sync, async, streaming, and batch APIs.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/langchain-request-coalescing",
            "instruction_sha256": "208b99162a6a2a40e25630f566c3b735d3fea91dbb3c7b3717f5a3a3c7ea3557",
            "task_toml_sha256": "efd60419b581db6e23c27d4c234b3af6bb4bdd976fa510e40ced09c7eaaaeb98",
            "language": "python",
            "category": "feature_request"
          },
          "repository": "https://github.com/langchain-ai/langchain",
          "base_revision": "7cef35b",
          "runs": [
            {
              "run_id": "tonight-20261012-followon/langchain-request-coalescing__9VF4pyY",
              "task_id": "langchain-request-coalescing",
              "campaign": "tonight-20261012-followon",
              "outcome": "graded-model-failure",
              "score": {
                "feature": {
                  "passed": 48,
                  "total": 50
                },
                "preservation": {
                  "passed": 232,
                  "total": 232
                },
                "reward": 0
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "agent": "OpenCode/Pier",
                "runtime": "Strata 0.1.42",
                "agent_timeout_seconds": 2700
              },
              "timing_seconds": {
                "agent_execution": 1811.456386,
                "environment_setup": 39.283825,
                "verifier": 42.620816
              },
              "usage": {
                "input_tokens": 8541308,
                "cached_tokens": 8462115,
                "output_tokens": 88554,
                "peak_context_tokens": 152623,
                "steps": 86,
                "requests": 87
              },
              "patch": {
                "bytes": 65957,
                "sha256": "163719b575919e86421221e347110045eca5ba083ad4a75bcaf5bd512674e0f3"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [
            {
              "model": "gpt-6-astra",
              "provider": "openai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_gpt_6_astra_xhigh",
              "reasoning_effort": "xhigh",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 2,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 3,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 4,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/langchain-request-coalescing",
              "source_html_sha256": "5958c99cc12f717bae4632f4ca4d41cc409157be76db71e3959736e7b221de47"
            },
            {
              "model": "claude-opus-5",
              "provider": "vertex_ai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_claude_opus_5_max",
              "reasoning_effort": "max",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/langchain-request-coalescing",
              "source_html_sha256": "5958c99cc12f717bae4632f4ca4d41cc409157be76db71e3959736e7b221de47"
            }
          ],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "mnamer-daemon-watch-lifecycle",
          "official_prompt": {
            "title": "Validate daemon watch, status, and log lifecycle",
            "description": "Add daemon watch validation, state tracking, logging, and lifecycle commands with non-blocking run-once processing.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/mnamer-daemon-watch-lifecycle",
            "instruction_sha256": "1759aa73299b2871d311613b6d53259f7b32dce806f6978542e5b111f5ab4649",
            "task_toml_sha256": "73b6024b768f0af0a178de6de7aa59ac460c2ec96fe1ca7cc9ac35269c118916",
            "language": "python",
            "category": "feature_request"
          },
          "repository": "https://github.com/jkwill87/mnamer",
          "base_revision": "73f5b537c8cad998e8e6d6bc40ad60e2e23bf268",
          "runs": [
            {
              "run_id": "tonight-20261012-followon/mnamer-daemon-watch-lifecycle__hjr3o3p",
              "task_id": "mnamer-daemon-watch-lifecycle",
              "campaign": "tonight-20261012-followon",
              "outcome": "pass",
              "score": {
                "feature": {
                  "passed": 51,
                  "total": 51
                },
                "preservation": {
                  "passed": 319,
                  "total": 319
                },
                "reward": 1
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "agent": "OpenCode/Pier",
                "runtime": "Strata 0.1.42",
                "agent_timeout_seconds": 2700
              },
              "timing_seconds": {
                "agent_execution": 842.466585,
                "environment_setup": 24.108183,
                "verifier": 36.395475
              },
              "usage": {
                "input_tokens": 2539459,
                "cached_tokens": 2489345,
                "output_tokens": 39188,
                "peak_context_tokens": 78953,
                "steps": 43,
                "requests": 44
              },
              "patch": {
                "bytes": 26441,
                "sha256": "c27e34c9f9b0b799056c7059ea227881456292232826185edf2c64a3dfd45628"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [
            {
              "model": "gpt-6-astra",
              "provider": "openai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_gpt_6_astra_xhigh",
              "reasoning_effort": "xhigh",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 2,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 3,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 4,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/mnamer-daemon-watch-lifecycle",
              "source_html_sha256": "b4e4ca798945141d762afbed1473f6483ffa611c9800a9e77bec9970bd61119a"
            },
            {
              "model": "claude-opus-5",
              "provider": "vertex_ai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_claude_opus_5_max",
              "reasoning_effort": "max",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/mnamer-daemon-watch-lifecycle",
              "source_html_sha256": "b4e4ca798945141d762afbed1473f6483ffa611c9800a9e77bec9970bd61119a"
            }
          ],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "mobly-grouped-test-barriers",
          "official_prompt": {
            "title": "Add grouped test phases with synchronized barriers",
            "description": "Add grouped execution phases with explicit synchronization barriers and per-group setup and teardown.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/mobly-grouped-test-barriers",
            "instruction_sha256": "475b2b59d25c8bf27ba72b9b079207cca38f1b39d250fb695cf1f47015a0f064",
            "task_toml_sha256": "872d09f2629bd0cb136f1e5e85191559641266f88477e48cb7c5c96af862545a",
            "language": "python",
            "category": "feature_request"
          },
          "repository": "https://github.com/google/mobly",
          "base_revision": "ec052921917ef201e73cc8e275dc91c5706b345f",
          "runs": [
            {
              "run_id": "tonight-20261012-followon/mobly-grouped-test-barriers__YfFtLYP",
              "task_id": "mobly-grouped-test-barriers",
              "campaign": "tonight-20261012-followon",
              "outcome": "graded-model-failure",
              "score": {
                "feature": {
                  "passed": 78,
                  "total": 79
                },
                "preservation": {
                  "passed": 808,
                  "total": 808
                },
                "reward": 0
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "agent": "OpenCode/Pier",
                "runtime": "Strata 0.1.42",
                "agent_timeout_seconds": 2700
              },
              "timing_seconds": {
                "agent_execution": 1142.854936,
                "environment_setup": 23.792344,
                "verifier": 19.916711
              },
              "usage": {
                "input_tokens": 5525553,
                "cached_tokens": 5483608,
                "output_tokens": 62903,
                "peak_context_tokens": 104517,
                "steps": 78,
                "requests": 79
              },
              "patch": {
                "bytes": 53296,
                "sha256": "edd622cf2b4cf3c3bf8d1d777d580f8a401568ed663172bfceb7e5f6e5f8188d"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [
            {
              "model": "gpt-6-astra",
              "provider": "openai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_gpt_6_astra_xhigh",
              "reasoning_effort": "xhigh",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 2,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 3,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 4,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/mobly-grouped-test-barriers",
              "source_html_sha256": "d2bc67239091cf839d913f30dc5e12eae97283b09d4a9194fbdc513a0f34bd3e"
            },
            {
              "model": "claude-opus-5",
              "provider": "vertex_ai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_claude_opus_5_max",
              "reasoning_effort": "max",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 2,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/mobly-grouped-test-barriers",
              "source_html_sha256": "d2bc67239091cf839d913f30dc5e12eae97283b09d4a9194fbdc513a0f34bd3e"
            }
          ],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "narwhals-rolling-window-suite",
          "official_prompt": {
            "title": "Add rolling min, max, median, and quantile methods",
            "description": "Add the remaining rolling window methods to Expr and Series with consistent validation and backend delegation.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/narwhals-rolling-window-suite",
            "instruction_sha256": "d25036c0761a9e6a2194f40bda5ae29b12487ee4f4c2045e4326ece0a80679a2",
            "task_toml_sha256": "d2b133d6a3a2d5d8d2f9474d34fcd3c5cd2f892dfc05326a6f4f7ceb169bd8b7",
            "language": "python",
            "category": "feature_request"
          },
          "repository": "https://github.com/narwhals-dev/narwhals",
          "base_revision": "061c97f8a01bf9e721835978b039303c5051501c",
          "runs": [
            {
              "run_id": "tonight-20261012-followon/narwhals-rolling-window-suite__LyjbpEx",
              "task_id": "narwhals-rolling-window-suite",
              "campaign": "tonight-20261012-followon",
              "outcome": "pass",
              "score": {
                "feature": {
                  "passed": 103,
                  "total": 103
                },
                "preservation": {
                  "passed": 10093,
                  "total": 10093
                },
                "reward": 1
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "agent": "OpenCode/Pier",
                "runtime": "Strata 0.1.42",
                "agent_timeout_seconds": 2700
              },
              "timing_seconds": {
                "agent_execution": 2281.09125,
                "environment_setup": 30.159255,
                "verifier": 82.595128
              },
              "usage": {
                "input_tokens": 14409783,
                "cached_tokens": 14179719,
                "output_tokens": 92623,
                "peak_context_tokens": 157740,
                "steps": 156,
                "requests": 157
              },
              "patch": {
                "bytes": 73788,
                "sha256": "53dc53defe7601d42f98fae8b14a4e31d58d2a66dd59f17866a0652a71f53c7e"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [
            {
              "model": "gpt-6-astra",
              "provider": "openai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_gpt_6_astra_xhigh",
              "reasoning_effort": "xhigh",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 2,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 3,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 4,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/narwhals-rolling-window-suite",
              "source_html_sha256": "403b9c797a992d2a232c0ba2c0eeb1ff5fb5985196b634be537c66ea981a25e6"
            },
            {
              "model": "claude-opus-5",
              "provider": "vertex_ai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_claude_opus_5_max",
              "reasoning_effort": "max",
              "attempts": 4,
              "passed": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": null,
                  "outcome": "pass",
                  "passed": true
                }
              ],
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/narwhals-rolling-window-suite",
              "source_html_sha256": "403b9c797a992d2a232c0ba2c0eeb1ff5fb5985196b634be537c66ea981a25e6"
            }
          ],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        }
      ]
    },
    {
      "id": "rails-pilot-20261012",
      "label": "Rails Lemans \u00b7 pilot \u00b7 2026-10-12",
      "adapter": "rails-lemans",
      "benchmark": "Rails AI Evals",
      "campaign": "rails-pilot-20261012",
      "benchmark_revision": "b799f091cf138bbc880baa6423eccd0d26066dbb",
      "harness_revision": "29994546b8f5c15e92984057f92ff5ff2c8061be",
      "native_score_semantics": "official verifier checks passed / checks collected",
      "tasks": [
        {
          "task_id": "ac-throttle-search",
          "official_prompt": {
            "title": "Throttle the public book search per address, with a counter that expires itself.",
            "description": "Anyone can search a published book from the search box on its page \u2014 no sign-in, and there never has been one. The request goes to `POST /books/:book_id/search`.",
            "description_source": "official task prompt",
            "url": "https://raw.githubusercontent.com/rails/ai-evals/b799f091cf138bbc880baa6423eccd0d26066dbb/tasks/ac-throttle-search/instruction.md",
            "instruction_sha256": "e53717c49038b333ed7cfa90b1cc6f2c314ba052e26143627e39325996b5db60"
          },
          "application": "Writebook",
          "benchmark_revision": "b799f091cf138bbc880baa6423eccd0d26066dbb",
          "runs": [
            {
              "run_id": "rails-pilot-20261012/ac-throttle-search",
              "task_id": "ac-throttle-search",
              "campaign": "rails-pilot-20261012",
              "outcome": "pass",
              "score": {
                "checks": {
                  "passed": 5,
                  "total": 5
                },
                "reward": 1
              },
              "configuration": {
                "max_tokens": 196608,
                "min_p": 0.0,
                "model": "openai/qwen3.8-flash-next-iq3_s#medium",
                "presence_penalty": 0.0,
                "reasoning_effort": "medium",
                "repetition_penalty": 1.0,
                "temperature": 1.0,
                "top_k": 20,
                "top_p": 0.95,
                "agent": "miniswen-installed",
                "lemans_version": "1.5.1"
              },
              "timing_seconds": {
                "agent_execution": 560.422,
                "wall_clock": 586.9
              },
              "usage": {
                "input_tokens": 1300103,
                "cached_tokens": 1268143,
                "output_tokens": 21243,
                "steps": 43
              },
              "patch": {
                "bytes": 3841,
                "sha256": "4b4d9732e2d7fa897c6e5b53f654a44a7a4a69d4b8011c48d807f00598abe4a8"
              },
              "verification": {
                "status": "recovered-grading",
                "verifier": "Rails Lemans official verifier",
                "evidence": "verified campaign report and native verifier checks",
                "regrade": {
                  "unchanged_patch": true,
                  "original_outcome": "infrastructure-invalid",
                  "regrade_evidence": "official verifier checks rerun on the unchanged submitted patch"
                }
              }
            }
          ],
          "frontier": []
        },
        {
          "task_id": "ar-atomic-import",
          "official_prompt": {
            "title": "All-or-nothing import: a guard return must not commit half the work.",
            "description": "Books can be filled from an export manifest through `Book#import_manifest`, which presses each entry as a section. Imports are supposed to be all-or-nothing \u2014 and support tickets say they are not: when a manifest has a bad entry partway through, the import reports failure, but every entry *before* the bad one is left sitting in the book. People re-run the fixed manifest and end up with the first half duplicated.",
            "description_source": "official task prompt",
            "url": "https://raw.githubusercontent.com/rails/ai-evals/b799f091cf138bbc880baa6423eccd0d26066dbb/tasks/ar-atomic-import/instruction.md",
            "instruction_sha256": "d33943c15428243ce106febaef5afd004f4fcf8f5f39a2a2a82c7aad6357a620"
          },
          "application": "Writebook",
          "benchmark_revision": "b799f091cf138bbc880baa6423eccd0d26066dbb",
          "runs": [
            {
              "run_id": "rails-pilot-20261012/ar-atomic-import",
              "task_id": "ar-atomic-import",
              "campaign": "rails-pilot-20261012",
              "outcome": "pass",
              "score": {
                "checks": {
                  "passed": 10,
                  "total": 10
                },
                "reward": 1
              },
              "configuration": {
                "max_tokens": 196608,
                "min_p": 0.0,
                "model": "openai/qwen3.8-flash-next-iq3_s#medium",
                "presence_penalty": 0.0,
                "reasoning_effort": "medium",
                "repetition_penalty": 1.0,
                "temperature": 1.0,
                "top_k": 20,
                "top_p": 0.95,
                "agent": "miniswen-installed",
                "lemans_version": "1.5.1"
              },
              "timing_seconds": {
                "agent_execution": 200.228,
                "wall_clock": 225.9
              },
              "usage": {
                "input_tokens": 130386,
                "cached_tokens": 123450,
                "output_tokens": 7205,
                "steps": 18
              },
              "patch": {
                "bytes": 6090,
                "sha256": "3e07084de5db125819415470337b4ad941d141ff43e513d42aadb0c751674674"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "Rails Lemans official verifier",
                "evidence": "verified campaign report and native verifier checks"
              }
            }
          ],
          "frontier": []
        },
        {
          "task_id": "ft-kamal-deploy",
          "official_prompt": {
            "title": "Give the open-source Fizzy a Kamal deployment for one box, with the deployment's settings and secrets outside the code.",
            "description": "We need to provide a config for developers to deploy Kamal for one box: a Linux server with Docker, SQLite, everything in one container, SSL from the proxy. Someone who forks the repo should fill in their hostname, their mail server and their secrets, run `kamal setup`, and have Fizzy at https://their-host with data that survives the next deploy. Background jobs run on the box.",
            "description_source": "official task prompt",
            "url": "https://raw.githubusercontent.com/rails/ai-evals/b799f091cf138bbc880baa6423eccd0d26066dbb/tasks/ft-kamal-deploy/instruction.md",
            "instruction_sha256": "b47fee71b55cc0f6eb1c30762ced2a1cda3b380a3d955768844ef8046ede15d2"
          },
          "application": "Fizzy",
          "benchmark_revision": "b799f091cf138bbc880baa6423eccd0d26066dbb",
          "runs": [
            {
              "run_id": "rails-pilot-20261012/ft-kamal-deploy",
              "task_id": "ft-kamal-deploy",
              "campaign": "rails-pilot-20261012",
              "outcome": "graded-model-failure",
              "score": {
                "checks": {
                  "passed": 3,
                  "total": 8
                },
                "reward": 0
              },
              "configuration": {
                "max_tokens": 196608,
                "min_p": 0.0,
                "model": "openai/qwen3.8-flash-next-iq3_s#medium",
                "presence_penalty": 0.0,
                "reasoning_effort": "medium",
                "repetition_penalty": 1.0,
                "temperature": 1.0,
                "top_k": 20,
                "top_p": 0.95,
                "agent": "miniswen-installed",
                "lemans_version": "1.5.1"
              },
              "timing_seconds": {
                "agent_execution": 698.126,
                "wall_clock": 745.0
              },
              "usage": {
                "input_tokens": 2490194,
                "cached_tokens": 2450996,
                "output_tokens": 31044,
                "steps": 62
              },
              "patch": {
                "bytes": 10961,
                "sha256": "fe35958bcee6377622e8fd28fc683d7a1017dbbb17e0e2dfa70fa3e80c7b9912"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "Rails Lemans official verifier",
                "evidence": "verified campaign report and native verifier checks"
              }
            }
          ],
          "frontier": []
        }
      ]
    },
    {
      "id": "rails-overnight-20261012",
      "label": "Rails Lemans \u00b7 overnight \u00b7 2026-10-12",
      "adapter": "rails-lemans",
      "benchmark": "Rails AI Evals",
      "campaign": "tonight-20261012/rails-runs",
      "benchmark_revision": "b799f091cf138bbc880baa6423eccd0d26066dbb",
      "harness_revision": "29994546b8f5c15e92984057f92ff5ff2c8061be",
      "native_score_semantics": "official verifier checks passed / checks collected",
      "tasks": [
        {
          "task_id": "aj-enqueue-after-commit",
          "official_prompt": {
            "title": "Hand an edit's follow-up jobs to the queue only once its transaction commits, without swallowing the notification an abandoned batch still owes its operator.",
            "description": "Whenever a book's cover changes, `Book::CoverProcessingJob` (in `app/jobs`) is kicked off from the `Book` model and stamps the cover's SHA-256 fingerprint into the blob's metadata. In production it fails intermittently: the worker picks it up before the database transaction that attached the cover has committed, finds no file, and raises `ActiveStorage::FileNotFoundError`. Worse, when a save inside a larger transaction is rolled back, the work for the never-committed cover is still handed off \u2014 and errors on every retry. The search entry a book edit refreshes goes the same way, off the same handover.",
            "description_source": "official task prompt",
            "url": "https://raw.githubusercontent.com/rails/ai-evals/b799f091cf138bbc880baa6423eccd0d26066dbb/tasks/aj-enqueue-after-commit/instruction.md",
            "instruction_sha256": "eb5cd52cdd406e2207345ce5518399c17d1e499a37a272d5062ea872b0ee3748"
          },
          "application": "Writebook",
          "benchmark_revision": "b799f091cf138bbc880baa6423eccd0d26066dbb",
          "runs": [
            {
              "run_id": "tonight-20261012/rails-runs/aj-enqueue-after-commit",
              "task_id": "aj-enqueue-after-commit",
              "campaign": "tonight-20261012/rails-runs",
              "outcome": "pass",
              "score": {
                "checks": {
                  "passed": 8,
                  "total": 8
                },
                "reward": 1
              },
              "configuration": {
                "max_tokens": 196608,
                "min_p": 0.0,
                "model": "openai/qwen3.8-flash-next-iq3_s#medium",
                "presence_penalty": 0.0,
                "reasoning_effort": "medium",
                "repetition_penalty": 1.0,
                "temperature": 1.0,
                "top_k": 20,
                "top_p": 0.95,
                "agent": "miniswen-installed",
                "lemans_version": "1.5.1"
              },
              "timing_seconds": {
                "agent_execution": 737.521,
                "wall_clock": 766.2
              },
              "usage": {
                "input_tokens": 1677007,
                "cached_tokens": 1646834,
                "output_tokens": 28487,
                "steps": 50
              },
              "patch": {
                "bytes": 6227,
                "sha256": "2157fc56f394fb7f81e3b193bbf9284d397bb18d64930152a5f4d93ddea0c63e"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "Rails Lemans official verifier",
                "evidence": "verified campaign report and native verifier checks"
              }
            }
          ],
          "frontier": []
        },
        {
          "task_id": "ar-tenant-isolation",
          "official_prompt": {
            "title": "Add tenants and close every cross-tenant read, including routes that never load a Book.",
            "description": "We're putting several customers on one Writebook install, so it needs tenants. A migration is expected this time: a `tenants` table with a name, a `tenant_id` on `users` and `books`, and every existing user and book swept into one tenant so running installs upgrade in place. New books land in their creator's tenant.",
            "description_source": "official task prompt",
            "url": "https://raw.githubusercontent.com/rails/ai-evals/b799f091cf138bbc880baa6423eccd0d26066dbb/tasks/ar-tenant-isolation/instruction.md",
            "instruction_sha256": "2bed05abd3ef03a46ff5fcf3e40f027b1b5c022858a0fa56bf6a4d999bef0a07"
          },
          "application": "Writebook",
          "benchmark_revision": "b799f091cf138bbc880baa6423eccd0d26066dbb",
          "runs": [
            {
              "run_id": "tonight-20261012/rails-runs/ar-tenant-isolation",
              "task_id": "ar-tenant-isolation",
              "campaign": "tonight-20261012/rails-runs",
              "outcome": "pass",
              "score": {
                "checks": {
                  "passed": 10,
                  "total": 10
                },
                "reward": 1
              },
              "configuration": {
                "max_tokens": 196608,
                "min_p": 0.0,
                "model": "openai/qwen3.8-flash-next-iq3_s#medium",
                "presence_penalty": 0.0,
                "reasoning_effort": "medium",
                "repetition_penalty": 1.0,
                "temperature": 1.0,
                "top_k": 20,
                "top_p": 0.95,
                "agent": "miniswen-installed",
                "lemans_version": "1.5.1"
              },
              "timing_seconds": {
                "agent_execution": 720.004,
                "wall_clock": 751.5
              },
              "usage": {
                "input_tokens": 1476840,
                "cached_tokens": 1446467,
                "output_tokens": 28849,
                "steps": 44
              },
              "patch": {
                "bytes": 12265,
                "sha256": "fba8e149be580fa674c9f2b5d7892368d6a883638cf22d75610e14b64b04d2bc"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "Rails Lemans official verifier",
                "evidence": "verified campaign report and native verifier checks"
              }
            }
          ],
          "frontier": []
        },
        {
          "task_id": "ft-card-reactions",
          "official_prompt": {
            "title": "Extend reaction support from comments to cards.",
            "description": "Currently, users can react only to comments within cards. Extend reaction support to cards themselves. Users should be able to react to a card directly from its detail view. Card previews in board columns should show the card's boost count when it is non-zero. In all other respects, card reactions should behave like comment reactions.",
            "description_source": "official task prompt",
            "url": "https://raw.githubusercontent.com/rails/ai-evals/b799f091cf138bbc880baa6423eccd0d26066dbb/tasks/ft-card-reactions/instruction.md",
            "instruction_sha256": "99bc5f8f6c946cacfad2e44a441175b87c17af8c83de6c4b94eaa54f0a22d8f9"
          },
          "application": "Fizzy",
          "benchmark_revision": "b799f091cf138bbc880baa6423eccd0d26066dbb",
          "runs": [
            {
              "run_id": "tonight-20261012/rails-runs/ft-card-reactions",
              "task_id": "ft-card-reactions",
              "campaign": "tonight-20261012/rails-runs",
              "outcome": "graded-model-failure",
              "score": {
                "checks": {
                  "passed": 6,
                  "total": 7
                },
                "reward": 0
              },
              "configuration": {
                "max_tokens": 196608,
                "min_p": 0.0,
                "model": "openai/qwen3.8-flash-next-iq3_s#medium",
                "presence_penalty": 0.0,
                "reasoning_effort": "medium",
                "repetition_penalty": 1.0,
                "temperature": 1.0,
                "top_k": 20,
                "top_p": 0.95,
                "agent": "miniswen-installed",
                "lemans_version": "1.5.1"
              },
              "timing_seconds": {
                "agent_execution": 1668.382,
                "wall_clock": 1711.3
              },
              "usage": {
                "input_tokens": 5372831,
                "cached_tokens": 5329245,
                "output_tokens": 53365,
                "steps": 106
              },
              "patch": {
                "bytes": 23662,
                "sha256": "d87d997e32a4d1f3ea3c2c09fc5364bd308d944ac388edcdd47a19f37fd853a7"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "Rails Lemans official verifier",
                "evidence": "verified campaign report and native verifier checks"
              }
            }
          ],
          "frontier": []
        }
      ]
    },
    {
      "id": "rails-lemans-tonight-20261013",
      "label": "Rails Lemans \u00b7 tonight-20261013",
      "adapter": "rails-lemans",
      "benchmark": "Rails AI Evals",
      "campaign": "tonight-20261013",
      "native_score_semantics": "official verifier checks passed / checks collected",
      "tasks": [
        {
          "task_id": "ac-deep-link-return",
          "official_prompt": {
            "title": "Open-redirect on sign-in return: reject hostile hosts, keep legitimate deep links.",
            "description": "Shared and emailed links carry the page they point at as `?return_to=`, and `SessionsController#create` sends the reader there once they sign in.",
            "description_source": "official task prompt",
            "url": "https://raw.githubusercontent.com/rails/ai-evals/b799f091cf138bbc880baa6423eccd0d26066dbb/tasks/ac-deep-link-return/instruction.md",
            "instruction_sha256": "8dbf17643aa7a581a0a26cd27dcf7b7684c28a11a3254a8ed8fb6e0d1f5fc5dd"
          },
          "application": "Writebook",
          "benchmark_revision": "b799f091cf138bbc880baa6423eccd0d26066dbb",
          "runs": [
            {
              "run_id": "rails-lemans-tonight-20261013/ac-deep-link-return",
              "task_id": "ac-deep-link-return",
              "campaign": "rails-lemans-tonight-20261013",
              "outcome": "graded-model-failure",
              "score": {
                "checks": {
                  "passed": 3,
                  "total": 4
                },
                "reward": 0
              },
              "configuration": {
                "max_tokens": 196608,
                "min_p": 0.0,
                "model": "openai/qwen3.8-flash-next-iq3_s#medium",
                "presence_penalty": 0.0,
                "reasoning_effort": "medium",
                "repetition_penalty": 1.0,
                "temperature": 1.0,
                "top_k": 20,
                "top_p": 0.95,
                "agent": "miniswen-installed",
                "lemans_version": "1.5.1"
              },
              "timing_seconds": {
                "agent_execution": 265.572,
                "wall_clock": 301.1
              },
              "usage": {
                "input_tokens": 231190,
                "cached_tokens": 221162,
                "output_tokens": 10601,
                "steps": 22
              },
              "patch": {
                "bytes": 3609,
                "sha256": "64b1b7546c91cbef785e843e5dcb9bca66633ef82007f2b8f6fbb40e79b23497"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "Rails Lemans official verifier",
                "evidence": "verified campaign report and native verifier checks"
              }
            }
          ],
          "frontier": []
        },
        {
          "task_id": "hw-scoped-broadcast",
          "official_prompt": {
            "title": "A Turbo broadcast still reaches collaborators whose access was revoked.",
            "description": "Writebook shows a live \"being edited\" indicator in a page's editor: while one person saves, everyone else with the page open sees who is editing. But a collaborator removed from the book keeps receiving those pushes: an editor left open while they still had access keeps showing who is editing \u2014 presence payloads, editor names included, for a page they can no longer edit. The payload must not go out on a channel a removed collaborator can still be listening on.",
            "description_source": "official task prompt",
            "url": "https://raw.githubusercontent.com/rails/ai-evals/b799f091cf138bbc880baa6423eccd0d26066dbb/tasks/hw-scoped-broadcast/instruction.md",
            "instruction_sha256": "e98b0badfcfb03fbb70cef6e945c0d7c237778e95e5a779f1a55a0e99567da30"
          },
          "application": "Writebook",
          "benchmark_revision": "b799f091cf138bbc880baa6423eccd0d26066dbb",
          "runs": [
            {
              "run_id": "rails-lemans-tonight-20261013/hw-scoped-broadcast",
              "task_id": "hw-scoped-broadcast",
              "campaign": "rails-lemans-tonight-20261013",
              "outcome": "graded-model-failure",
              "score": {
                "checks": {
                  "passed": 7,
                  "total": 8
                },
                "reward": 0
              },
              "configuration": {
                "max_tokens": 196608,
                "min_p": 0.0,
                "model": "openai/qwen3.8-flash-next-iq3_s#medium",
                "presence_penalty": 0.0,
                "reasoning_effort": "medium",
                "repetition_penalty": 1.0,
                "temperature": 1.0,
                "top_k": 20,
                "top_p": 0.95,
                "agent": "miniswen-installed",
                "lemans_version": "1.5.1"
              },
              "timing_seconds": {
                "agent_execution": 409.777,
                "wall_clock": 448.5
              },
              "usage": {
                "input_tokens": 571670,
                "cached_tokens": 553184,
                "output_tokens": 17173,
                "steps": 30
              },
              "patch": {
                "bytes": 4170,
                "sha256": "79d4741ce4dedeb376f55c11f30951eceec1af10673778dc25120d071054b7a7"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "Rails Lemans official verifier",
                "evidence": "verified campaign report and native verifier checks"
              }
            }
          ],
          "frontier": []
        },
        {
          "task_id": "ft-activity-feed-api",
          "official_prompt": {
            "title": "Expose the existing HTML activity timeline as an API endpoint.",
            "description": "Cards, comments, boards, and other resources are available through the API, but account activity is only available in the HTML timeline. Add `GET /:account_slug/activities.json` so API clients can consume account activity directly.",
            "description_source": "official task prompt",
            "url": "https://raw.githubusercontent.com/rails/ai-evals/b799f091cf138bbc880baa6423eccd0d26066dbb/tasks/ft-activity-feed-api/instruction.md",
            "instruction_sha256": "92128ed45be447fa675111f11f9bd051f394af3eb8a8e75837f3fc55d616b06b"
          },
          "application": "Fizzy",
          "benchmark_revision": "b799f091cf138bbc880baa6423eccd0d26066dbb",
          "runs": [
            {
              "run_id": "rails-lemans-tonight-20261013/ft-activity-feed-api",
              "task_id": "ft-activity-feed-api",
              "campaign": "rails-lemans-tonight-20261013",
              "outcome": "graded-model-failure",
              "score": {
                "checks": {
                  "passed": 13,
                  "total": 14
                },
                "reward": 0
              },
              "configuration": {
                "max_tokens": 196608,
                "min_p": 0.0,
                "model": "openai/qwen3.8-flash-next-iq3_s#medium",
                "presence_penalty": 0.0,
                "reasoning_effort": "medium",
                "repetition_penalty": 1.0,
                "temperature": 1.0,
                "top_k": 20,
                "top_p": 0.95,
                "agent": "miniswen-installed",
                "lemans_version": "1.5.1"
              },
              "timing_seconds": {
                "agent_execution": 661.428,
                "wall_clock": 707.7
              },
              "usage": {
                "input_tokens": 2072301,
                "cached_tokens": 2030766,
                "output_tokens": 24691,
                "steps": 54
              },
              "patch": {
                "bytes": 8912,
                "sha256": "b483fa85c73d76cab6de9fc4730f5e76ec0a84b310e941cbf5d0f67990b15dd5"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "Rails Lemans official verifier",
                "evidence": "verified campaign report and native verifier checks"
              }
            }
          ],
          "frontier": []
        },
        {
          "task_id": "ft-webhook-delivery-trust-and-restart",
          "official_prompt": {
            "title": "Make board webhooks deliver, once per event, switched off when dead.",
            "description": "Add webhooks for boards, on top of the existing webhook settings. When an event happens, the receiver gets a POST about it and the log shows how it went. Deliver exactly once.",
            "description_source": "official task prompt",
            "url": "https://raw.githubusercontent.com/rails/ai-evals/b799f091cf138bbc880baa6423eccd0d26066dbb/tasks/ft-webhook-delivery-trust-and-restart/instruction.md",
            "instruction_sha256": "f5a7585e3a10313a55812c47fb1942dc67c6a7583d8fa975415ac66319c3151e"
          },
          "application": "Fizzy",
          "benchmark_revision": "b799f091cf138bbc880baa6423eccd0d26066dbb",
          "runs": [
            {
              "run_id": "rails-lemans-tonight-20261013/ft-webhook-delivery-trust-and-restart",
              "task_id": "ft-webhook-delivery-trust-and-restart",
              "campaign": "rails-lemans-tonight-20261013",
              "outcome": "graded-model-failure",
              "score": {
                "checks": {
                  "passed": 6,
                  "total": 9
                },
                "reward": 0
              },
              "configuration": {
                "max_tokens": 196608,
                "min_p": 0.0,
                "model": "openai/qwen3.8-flash-next-iq3_s#medium",
                "presence_penalty": 0.0,
                "reasoning_effort": "medium",
                "repetition_penalty": 1.0,
                "temperature": 1.0,
                "top_k": 20,
                "top_p": 0.95,
                "agent": "miniswen-installed",
                "lemans_version": "1.5.1"
              },
              "timing_seconds": {
                "agent_execution": 1261.533,
                "wall_clock": 1307.2
              },
              "usage": {
                "input_tokens": 4243255,
                "cached_tokens": 4184557,
                "output_tokens": 53891,
                "steps": 70
              },
              "patch": {
                "bytes": 21862,
                "sha256": "92154fb1e61256d6f7e94475c139805c308c00ad14a4220e2c573d73eb9192d5"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "Rails Lemans official verifier",
                "evidence": "verified campaign report and native verifier checks"
              }
            }
          ],
          "frontier": []
        },
        {
          "task_id": "ft-board-publish-unpublish-public-boundary",
          "official_prompt": {
            "title": "Publish a board to an unguessable public URL that unpublishing revokes at once.",
            "description": "We want to make it possible to publish boards, so anyone with a secret public URL (`/public/boards/:key`) can see them. Only admins can publish/unpublish boards (`POST`/`DELETE` on `/boards/:id/publication`).",
            "description_source": "official task prompt",
            "url": "https://raw.githubusercontent.com/rails/ai-evals/b799f091cf138bbc880baa6423eccd0d26066dbb/tasks/ft-board-publish-unpublish-public-boundary/instruction.md",
            "instruction_sha256": "a6cad41844f5d668e5b428553f1e4a19e806531747b8d7db6bed33dc9be7aadd"
          },
          "application": "Fizzy",
          "benchmark_revision": "b799f091cf138bbc880baa6423eccd0d26066dbb",
          "runs": [
            {
              "run_id": "rails-lemans-tonight-20261013/ft-board-publish-unpublish-public-boundary",
              "task_id": "ft-board-publish-unpublish-public-boundary",
              "campaign": "rails-lemans-tonight-20261013",
              "outcome": "graded-model-failure",
              "score": {
                "checks": {
                  "passed": 10,
                  "total": 12
                },
                "reward": 0
              },
              "configuration": {
                "max_tokens": 196608,
                "min_p": 0.0,
                "model": "openai/qwen3.8-flash-next-iq3_s#medium",
                "presence_penalty": 0.0,
                "reasoning_effort": "medium",
                "repetition_penalty": 1.0,
                "temperature": 1.0,
                "top_k": 20,
                "top_p": 0.95,
                "agent": "miniswen-installed",
                "lemans_version": "1.5.1"
              },
              "timing_seconds": {
                "agent_execution": 996.843,
                "wall_clock": 1046.0
              },
              "usage": {
                "input_tokens": 4275027,
                "cached_tokens": 4210304,
                "output_tokens": 41116,
                "steps": 69
              },
              "patch": {
                "bytes": 20010,
                "sha256": "1b3ab2c3e26ff5efb2ca850f07863ab839be57eb10bd41a443c25f4a3f498af8"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "Rails Lemans official verifier",
                "evidence": "verified campaign report and native verifier checks"
              }
            }
          ],
          "frontier": []
        }
      ],
      "benchmark_revision": "b799f091cf138bbc880baa6423eccd0d26066dbb",
      "harness_revision": "29994546b8f5c15e92984057f92ff5ff2c8061be"
    },
    {
      "id": "deepswe-tonight-20261013",
      "label": "DeepSWE \u00b7 tonight-20261013",
      "adapter": "deepswe",
      "benchmark": "DeepSWE v1.1",
      "campaign": "tonight-20261013",
      "native_score_semantics": "feature-to-patch and patch-to-patch verifier checks",
      "tasks": [
        {
          "task_id": "pest-character-class-coalescing",
          "official_prompt": {
            "title": "Coalesce qualifying choices into character classes",
            "description": "Add optimizer passes that collapse qualifying choice chains into merged character and negated character classes.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/pest-character-class-coalescing",
            "instruction_sha256": "0368afdee566cf6b3685fb3be6432afb26104661c06c8d3102175eb8c2a31438",
            "task_toml_sha256": "f27e32038560083ffc87d25564ef695ccda4ee18e8b4d82ae3766559d68ed1b0",
            "language": "rust",
            "category": "feature_request"
          },
          "repository": "https://github.com/pest-parser/pest",
          "base_revision": "79dd30d11aab6f0fba3cd79bd48f456209b966b3",
          "runs": [
            {
              "run_id": "tonight-20261013/pest-character-class-coalescing__5mURYHS",
              "task_id": "pest-character-class-coalescing",
              "campaign": "tonight-20261013",
              "outcome": "graded-model-failure",
              "score": {
                "feature": {
                  "passed": 99,
                  "total": 104
                },
                "preservation": {
                  "passed": 249,
                  "total": 250
                },
                "reward": 0
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "agent": "OpenCode/Pier",
                "agent_version": "1.18.31",
                "context_tokens": 196608,
                "agent_timeout_seconds": 2700.0,
                "compaction_reserved_tokens": 40000
              },
              "timing_seconds": {
                "environment_setup": 37.471339,
                "agent_setup": 0.277941,
                "agent_execution": 739.710135,
                "verifier": 25.345051
              },
              "usage": {
                "input_tokens": 3201389,
                "cached_tokens": 3161060,
                "output_tokens": 35082,
                "peak_context_tokens": 74145,
                "steps": 64,
                "requests": 65
              },
              "patch": {
                "bytes": 23543,
                "sha256": "a48a8c573b06ec88260c9ac119e9e3f9cdc0fe1ddc18c60e32508f28f9e60db6"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [
            {
              "model": "gpt-6-astra",
              "provider": "openai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_gpt_6_astra_xhigh",
              "reasoning_effort": "xhigh",
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/pest-character-class-coalescing",
              "passed": 0,
              "attempts": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 2,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 3,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 4,
                  "outcome": "fail",
                  "passed": false
                }
              ]
            },
            {
              "model": "claude-opus-5",
              "provider": "vertex_ai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_claude_opus_5_max",
              "reasoning_effort": "max",
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/pest-character-class-coalescing",
              "passed": 4,
              "attempts": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 2,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 3,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 4,
                  "outcome": "pass",
                  "passed": true
                }
              ]
            }
          ],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "ink-grid-box-layout",
          "official_prompt": {
            "title": "Add CSS Grid layout to the Box component",
            "description": "Add CSS Grid layout parsing and placement to the Box component, including track sizing, gaps, and explicit child positioning.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/ink-grid-box-layout",
            "instruction_sha256": "8913c555865b03f915888ad262db805a5d6b004b01b034c6502fe64a5938a34b",
            "task_toml_sha256": "c75064d6aa2d3a136ef42d6f56257d32f992eab663caf182c4d17adc4a4e18e8",
            "language": "typescript",
            "category": "feature_request"
          },
          "repository": "https://github.com/vadimdemedes/ink",
          "base_revision": "0cea59169ef0f3f83e4aa7fbedbff9d165646472",
          "runs": [
            {
              "run_id": "tonight-20261013/ink-grid-box-layout__S8teBWM",
              "task_id": "ink-grid-box-layout",
              "campaign": "tonight-20261013",
              "outcome": "graded-model-failure",
              "score": {
                "feature": {
                  "passed": 17,
                  "total": 25
                },
                "preservation": {
                  "passed": 49,
                  "total": 49
                },
                "reward": 0
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "agent": "OpenCode/Pier",
                "agent_version": "1.18.31",
                "context_tokens": 196608,
                "agent_timeout_seconds": 2700.0,
                "compaction_reserved_tokens": 40000
              },
              "timing_seconds": {
                "environment_setup": 29.540605,
                "agent_setup": 0.293095,
                "agent_execution": 1135.449174,
                "verifier": 22.49806
              },
              "usage": {
                "input_tokens": 3386444,
                "cached_tokens": 3349240,
                "output_tokens": 39375,
                "peak_context_tokens": 72413,
                "steps": 66,
                "requests": 67
              },
              "patch": {
                "bytes": 23613,
                "sha256": "dd188b2ac3b51180d655ff07c316a22513ca46be12e301927416cc8aeea5f92a"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [
            {
              "model": "gpt-6-astra",
              "provider": "openai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_gpt_6_astra_xhigh",
              "reasoning_effort": "xhigh",
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/ink-grid-box-layout",
              "passed": 0,
              "attempts": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 2,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 3,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 4,
                  "outcome": "fail",
                  "passed": false
                }
              ]
            },
            {
              "model": "claude-opus-5",
              "provider": "vertex_ai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_claude_opus_5_max",
              "reasoning_effort": "max",
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/ink-grid-box-layout",
              "passed": 0,
              "attempts": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 2,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 3,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 4,
                  "outcome": "fail",
                  "passed": false
                }
              ]
            }
          ],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "meriyah-explicit-resource-declarations",
          "official_prompt": {
            "title": "Add explicit resource management declarations to the parser",
            "description": "Add parsing and AST support for `using` and `await using` declarations, including context-sensitive errors.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/meriyah-explicit-resource-declarations",
            "instruction_sha256": "a85dfeb98a4203f9eb7392f918543fab27d83f344a92d290454aedc145c7b5f7",
            "task_toml_sha256": "1c245857d89c6c4f0d12922bd83d0a2eeeb56277e2d9c01094a0421daa2af905",
            "language": "typescript",
            "category": "feature_request"
          },
          "repository": "https://github.com/meriyah/meriyah",
          "base_revision": "d141eb14a40b79c04d1b1db5c20c6afa3844c0d9",
          "runs": [
            {
              "run_id": "tonight-20261013/meriyah-explicit-resource-declar__R6xK8ck",
              "task_id": "meriyah-explicit-resource-declarations",
              "campaign": "tonight-20261013",
              "outcome": "graded-model-failure",
              "score": {
                "feature": {
                  "passed": 42,
                  "total": 49
                },
                "preservation": {
                  "passed": 51469,
                  "total": 51469
                },
                "reward": 0
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "agent": "OpenCode/Pier",
                "agent_version": "1.18.31",
                "context_tokens": 196608,
                "agent_timeout_seconds": 2700.0,
                "compaction_reserved_tokens": 40000
              },
              "timing_seconds": {
                "environment_setup": 33.156144,
                "agent_setup": 0.281629,
                "agent_execution": 1435.542323,
                "verifier": 60.060344
              },
              "usage": {
                "input_tokens": 8755813,
                "cached_tokens": 8682690,
                "output_tokens": 65902,
                "peak_context_tokens": 132420,
                "steps": 107,
                "requests": 108
              },
              "patch": {
                "bytes": 43397,
                "sha256": "b27bf084db46945eb94bdac363711a682bf5204fa5e7ba3bd7c499b00703d261"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [
            {
              "model": "gpt-6-astra",
              "provider": "openai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_gpt_6_astra_xhigh",
              "reasoning_effort": "xhigh",
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/meriyah-explicit-resource-declarations",
              "passed": 0,
              "attempts": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 2,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 3,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 4,
                  "outcome": "fail",
                  "passed": false
                }
              ]
            },
            {
              "model": "claude-opus-5",
              "provider": "vertex_ai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_claude_opus_5_max",
              "reasoning_effort": "max",
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/meriyah-explicit-resource-declarations",
              "passed": 0,
              "attempts": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 2,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 3,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 4,
                  "outcome": "fail",
                  "passed": false
                }
              ]
            }
          ],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "go-critic-doc-link-checker",
          "official_prompt": {
            "title": "Add a checker for broken doc comment links",
            "description": "Add a diagnostic checker that validates doc comment symbol links against package and type information.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/go-critic-doc-link-checker",
            "instruction_sha256": "101d1e82dfbb0514de914d8be1912f99f74a3f9c450ea8816cf3443a7faf24fb",
            "task_toml_sha256": "38ca3fb39ac476817895615ce3cf962c223e57e0e2359b64d16f19aa457b747e",
            "language": "go",
            "category": "feature_request"
          },
          "repository": "https://github.com/go-critic/go-critic",
          "base_revision": "9aea378c4dccd6f4394196ad8f0873b3e84678c8",
          "runs": [
            {
              "run_id": "tonight-20261013/go-critic-doc-link-checker__iRfrCsy",
              "task_id": "go-critic-doc-link-checker",
              "campaign": "tonight-20261013",
              "outcome": "graded-model-failure",
              "score": {
                "feature": {
                  "passed": 2,
                  "total": 3
                },
                "preservation": {
                  "passed": 15,
                  "total": 16
                },
                "reward": 0
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "agent": "OpenCode/Pier",
                "agent_version": "1.18.31",
                "context_tokens": 196608,
                "agent_timeout_seconds": 2700.0,
                "compaction_reserved_tokens": 40000
              },
              "timing_seconds": {
                "environment_setup": 26.020946,
                "agent_setup": 0.294206,
                "agent_execution": 1417.433721,
                "verifier": 30.635816
              },
              "usage": {
                "input_tokens": 4661185,
                "cached_tokens": 4599518,
                "output_tokens": 63503,
                "peak_context_tokens": 102402,
                "steps": 78,
                "requests": 79
              },
              "patch": {
                "bytes": 19792,
                "sha256": "d8ba35864708e41d7657f5e3cbe7fd212144f02326fb826af44cbae475b54e59"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [
            {
              "model": "gpt-6-astra",
              "provider": "openai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_gpt_6_astra_xhigh",
              "reasoning_effort": "xhigh",
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/go-critic-doc-link-checker",
              "passed": 0,
              "attempts": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 2,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 3,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 4,
                  "outcome": "fail",
                  "passed": false
                }
              ]
            },
            {
              "model": "claude-opus-5",
              "provider": "vertex_ai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_claude_opus_5_max",
              "reasoning_effort": "max",
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/go-critic-doc-link-checker",
              "passed": 4,
              "attempts": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 2,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 3,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 4,
                  "outcome": "pass",
                  "passed": true
                }
              ]
            }
          ],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "sqlfmt-create-table-ddl-formatting",
          "official_prompt": {
            "title": "Format CREATE TABLE DDL and add DDL parsing helpers",
            "description": "Format CREATE TABLE statements with DDL-aware line breaking and add parsing models for table columns and constraints.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/sqlfmt-create-table-ddl-formatting",
            "instruction_sha256": "9e445e5be9287df6e9843c4d630fc4ddd268775b4511a04f8e6560933064693c",
            "task_toml_sha256": "e915832661a266670f82bf9a3b3f0c589525ab822d6252acecdd0da0c8c98df3",
            "language": "python",
            "category": "feature_request"
          },
          "repository": "https://github.com/tconbeer/sqlfmt",
          "base_revision": "da140993a4547170ef85dc5ce7ce1c270f4322b3",
          "runs": [
            {
              "run_id": "tonight-20261013/sqlfmt-create-table-ddl-formatti__GhFpFuF",
              "task_id": "sqlfmt-create-table-ddl-formatting",
              "campaign": "tonight-20261013",
              "outcome": "graded-model-failure",
              "score": {
                "feature": {
                  "passed": 30,
                  "total": 32
                },
                "preservation": {
                  "passed": 1241,
                  "total": 1273
                },
                "reward": 0
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "agent": "OpenCode/Pier",
                "agent_version": "1.18.31",
                "context_tokens": 196608,
                "agent_timeout_seconds": 2700.0,
                "compaction_reserved_tokens": 40000
              },
              "timing_seconds": {
                "environment_setup": 25.248353,
                "agent_setup": 0.295324,
                "agent_execution": 2042.657127,
                "verifier": 41.359024
              },
              "usage": {
                "input_tokens": 8449806,
                "cached_tokens": 8208777,
                "output_tokens": 96804,
                "peak_context_tokens": 155629,
                "steps": 86,
                "requests": 87
              },
              "patch": {
                "bytes": 42254,
                "sha256": "6538b997e60aa52c4604892990ccd11bc2c8e70b33f9bf282f483be50ee8ee31"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [
            {
              "model": "gpt-6-astra",
              "provider": "openai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_gpt_6_astra_xhigh",
              "reasoning_effort": "xhigh",
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/sqlfmt-create-table-ddl-formatting",
              "passed": 0,
              "attempts": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 2,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 3,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 4,
                  "outcome": "fail",
                  "passed": false
                }
              ]
            },
            {
              "model": "claude-opus-5",
              "provider": "vertex_ai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_claude_opus_5_max",
              "reasoning_effort": "max",
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/sqlfmt-create-table-ddl-formatting",
              "passed": 0,
              "attempts": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 2,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 3,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 4,
                  "outcome": "fail",
                  "passed": false
                }
              ]
            }
          ],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "termenv-preserve-ansi-resets",
          "official_prompt": {
            "title": "Preserve ANSI resets during truncation and styling",
            "description": "Add ANSI tokenization, reset-preserving truncation, and style-aware output helpers.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/termenv-preserve-ansi-resets",
            "instruction_sha256": "a48484e41aedb6ab3ad0bc6deff51983b4ed47302ceea642e03d9137ae9ee421",
            "task_toml_sha256": "8e0ca91dabb5a1b8fc4ce62bdc3c1ac9f1711aea6e327d80d859361f67382533",
            "language": "go",
            "category": "feature_request"
          },
          "repository": "https://github.com/muesli/termenv",
          "base_revision": "368a3572b8146cc038b3f240da6792003d7e42c5",
          "runs": [
            {
              "run_id": "tonight-20261013/termenv-preserve-ansi-resets__hQfwhVG",
              "task_id": "termenv-preserve-ansi-resets",
              "campaign": "tonight-20261013",
              "outcome": "graded-model-failure",
              "score": {
                "feature": {
                  "passed": 31,
                  "total": 35
                },
                "preservation": {
                  "passed": 87,
                  "total": 87
                },
                "reward": 0
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "agent": "OpenCode/Pier",
                "agent_version": "1.18.31",
                "context_tokens": 196608,
                "agent_timeout_seconds": 2700.0,
                "compaction_reserved_tokens": 40000
              },
              "timing_seconds": {
                "environment_setup": 25.997532,
                "agent_setup": 0.275822,
                "agent_execution": 794.432187,
                "verifier": 24.492206
              },
              "usage": {
                "input_tokens": 1714454,
                "cached_tokens": 1673570,
                "output_tokens": 45595,
                "peak_context_tokens": 67937,
                "steps": 37,
                "requests": 38
              },
              "patch": {
                "bytes": 33466,
                "sha256": "b2b27c2b20eb41449c334dfbab0ed8e3211fe6f922545a6aebcd00e88e8c2549"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [
            {
              "model": "gpt-6-astra",
              "provider": "openai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_gpt_6_astra_xhigh",
              "reasoning_effort": "xhigh",
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/termenv-preserve-ansi-resets",
              "passed": 0,
              "attempts": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 2,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 3,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 4,
                  "outcome": "fail",
                  "passed": false
                }
              ]
            },
            {
              "model": "claude-opus-5",
              "provider": "vertex_ai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_claude_opus_5_max",
              "reasoning_effort": "max",
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/termenv-preserve-ansi-resets",
              "passed": 0,
              "attempts": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 2,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 3,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 4,
                  "outcome": "fail",
                  "passed": false
                }
              ]
            }
          ],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "helm-array-merge-strategies",
          "official_prompt": {
            "title": "Add configurable array merge strategies to Helm value coalescing",
            "description": "Add chart-scoped append and key-based merge strategies for coalescing arrays during Helm value merging.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/helm-array-merge-strategies",
            "instruction_sha256": "4d39c1963676acb1c1cdab6ac09d91fd5c321a77939b2db0cbbb666c911e47f6",
            "task_toml_sha256": "ce111f2acc7237a7c1e767fcb871a91ccdeea767bfb87f6f21c058efbb530626",
            "language": "go",
            "category": "feature_request"
          },
          "repository": "https://github.com/helm/helm",
          "base_revision": "42f78ba60edf531d5161e00d9819a7c34d976343",
          "runs": [
            {
              "run_id": "tonight-20261013/helm-array-merge-strategies__J4oi2S8",
              "task_id": "helm-array-merge-strategies",
              "campaign": "tonight-20261013",
              "outcome": "graded-model-failure",
              "score": {
                "feature": {
                  "passed": 47,
                  "total": 47
                },
                "preservation": {
                  "passed": 10,
                  "total": 12
                },
                "reward": 0
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "agent": "OpenCode/Pier",
                "agent_version": "1.18.31",
                "context_tokens": 196608,
                "agent_timeout_seconds": 2700.0,
                "compaction_reserved_tokens": 40000
              },
              "timing_seconds": {
                "environment_setup": 25.546308,
                "agent_setup": 0.292533,
                "agent_execution": 1785.979937,
                "verifier": 102.327377
              },
              "usage": {
                "input_tokens": 10897991,
                "cached_tokens": 10821878,
                "output_tokens": 75689,
                "peak_context_tokens": 147452,
                "steps": 112,
                "requests": 113
              },
              "patch": {
                "bytes": 64644,
                "sha256": "281cd634c7be14e3e0d7d845b3d173dd04b0a7e4cb46f05c25753bade03d2a53"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [
            {
              "model": "gpt-6-astra",
              "provider": "openai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_gpt_6_astra_xhigh",
              "reasoning_effort": "xhigh",
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/helm-array-merge-strategies",
              "passed": 0,
              "attempts": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 2,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 3,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 4,
                  "outcome": "fail",
                  "passed": false
                }
              ]
            },
            {
              "model": "claude-opus-5",
              "provider": "vertex_ai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_claude_opus_5_max",
              "reasoning_effort": "max",
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/helm-array-merge-strategies",
              "passed": 2,
              "attempts": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 2,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 3,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 4,
                  "outcome": "pass",
                  "passed": true
                }
              ]
            }
          ],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "happy-dom-deterministic-intersectionobserver",
          "official_prompt": {
            "title": "Implement a deterministic IntersectionObserver in Happy DOM",
            "description": "Implement a real, deterministic IntersectionObserver with async delivery, thresholds, root margins, and target tracking.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/happy-dom-deterministic-intersectionobserver",
            "instruction_sha256": "c294cb6d4e31f4c274fe79c373afa7d79ce36e9839739d7bac899fb6839ac368",
            "task_toml_sha256": "9a590906c60ed3bac08f96331cdd719f781f7e72a55a6fa82648f966acb25409",
            "language": "typescript",
            "category": "feature_request"
          },
          "repository": "https://github.com/capricorn86/happy-dom",
          "base_revision": "82a0888cb2c87a6123e05424b528f8e8c9b3e426",
          "runs": [
            {
              "run_id": "tonight-20261013/happy-dom-deterministic-intersec__nn2y6ja",
              "task_id": "happy-dom-deterministic-intersectionobserver",
              "campaign": "tonight-20261013",
              "outcome": "graded-model-failure",
              "score": {
                "feature": {
                  "passed": 13,
                  "total": 14
                },
                "preservation": {
                  "passed": 9,
                  "total": 9
                },
                "reward": 0
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "agent": "OpenCode/Pier",
                "agent_version": "1.18.31",
                "context_tokens": 196608,
                "agent_timeout_seconds": 2700.0,
                "compaction_reserved_tokens": 40000
              },
              "timing_seconds": {
                "environment_setup": 24.158872,
                "agent_setup": 0.278709,
                "agent_execution": 751.129845,
                "verifier": 21.697702
              },
              "usage": {
                "input_tokens": 1719320,
                "cached_tokens": 1693071,
                "output_tokens": 32818,
                "peak_context_tokens": 58768,
                "steps": 48,
                "requests": 49
              },
              "patch": {
                "bytes": 37432,
                "sha256": "024231386accd4c7a1eb008c2a4ac6bd354be850912e2f96d0456f6be9336c09"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [
            {
              "model": "gpt-6-astra",
              "provider": "openai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_gpt_6_astra_xhigh",
              "reasoning_effort": "xhigh",
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/happy-dom-deterministic-intersectionobserver",
              "passed": 0,
              "attempts": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 2,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 3,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 4,
                  "outcome": "fail",
                  "passed": false
                }
              ]
            },
            {
              "model": "claude-opus-5",
              "provider": "vertex_ai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_claude_opus_5_max",
              "reasoning_effort": "max",
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/happy-dom-deterministic-intersectionobserver",
              "passed": 2,
              "attempts": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 2,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 3,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 4,
                  "outcome": "pass",
                  "passed": true
                }
              ]
            }
          ],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "igel-persist-feature-schema",
          "official_prompt": {
            "title": "Persist the fitted feature schema across evaluate, predict, serve, and export",
            "description": "Persist and reuse the fitted feature schema so evaluation, prediction, serving, and export all apply the same canonicalized inputs.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/igel-persist-feature-schema",
            "instruction_sha256": "af7e38939072f50ed4d0fa9743dce1d963140700d4623289fa7ecfbcae301c49",
            "task_toml_sha256": "f6bfad291ade307db9897cc4f27e00b01604eca70a4bd8037359c7bb58070740",
            "language": "python",
            "category": "feature_request"
          },
          "repository": "https://github.com/nidhaloff/igel",
          "base_revision": "bf4544d6c86ab4ace21254cb38a011ce3e845700",
          "runs": [
            {
              "run_id": "tonight-20261013/igel-persist-feature-schema__YC6ku9M",
              "task_id": "igel-persist-feature-schema",
              "campaign": "tonight-20261013",
              "outcome": "pass",
              "score": {
                "feature": {
                  "passed": 24,
                  "total": 24
                },
                "preservation": {
                  "passed": 2,
                  "total": 2
                },
                "reward": 1
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "agent": "OpenCode/Pier",
                "agent_version": "1.18.31",
                "context_tokens": 196608,
                "agent_timeout_seconds": 2700.0,
                "compaction_reserved_tokens": 40000
              },
              "timing_seconds": {
                "environment_setup": 37.60312,
                "agent_setup": 0.28336,
                "agent_execution": 781.424692,
                "verifier": 20.99511
              },
              "usage": {
                "input_tokens": 3684078,
                "cached_tokens": 3644045,
                "output_tokens": 39369,
                "peak_context_tokens": 76989,
                "steps": 69,
                "requests": 70
              },
              "patch": {
                "bytes": 31831,
                "sha256": "7a66bda375fa7173385c1af686e41d9c361983b84959c309b02c6ba3aa954418"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [
            {
              "model": "gpt-6-astra",
              "provider": "openai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_gpt_6_astra_xhigh",
              "reasoning_effort": "xhigh",
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/igel-persist-feature-schema",
              "passed": 0,
              "attempts": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 2,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 3,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 4,
                  "outcome": "fail",
                  "passed": false
                }
              ]
            },
            {
              "model": "claude-opus-5",
              "provider": "vertex_ai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_claude_opus_5_max",
              "reasoning_effort": "max",
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/igel-persist-feature-schema",
              "passed": 2,
              "attempts": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 2,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 3,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 4,
                  "outcome": "fail",
                  "passed": false
                }
              ]
            }
          ],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "awilix-async-container-initialization",
          "official_prompt": {
            "title": "Add dependency-aware async initialization to the container",
            "description": "Add dependency-aware asynchronous container initialization with ordered startup, concurrency limits, and rollback on failure.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/awilix-async-container-initialization",
            "instruction_sha256": "340a292a209ce1ad90dc30ba47d0062f543805cf9e212bce662b0fb6bd5d3c66",
            "task_toml_sha256": "051f78a9f7c1641dfa9a362dacf26a8ea3b4100eaade210db2a74ef23061fc3d",
            "language": "typescript",
            "category": "feature_request"
          },
          "repository": "https://github.com/jeffijoe/awilix",
          "base_revision": "82ac179c1de4c216c4e333093044fac643303f0c",
          "runs": [
            {
              "run_id": "tonight-20261013/awilix-async-container-initializ__ZGFWNPp",
              "task_id": "awilix-async-container-initialization",
              "campaign": "tonight-20261013",
              "outcome": "graded-model-failure",
              "score": {
                "feature": {
                  "passed": 23,
                  "total": 24
                },
                "preservation": {
                  "passed": 162,
                  "total": 162
                },
                "reward": 0
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "agent": "OpenCode/Pier",
                "agent_version": "1.18.31",
                "context_tokens": 196608,
                "agent_timeout_seconds": 2700.0,
                "compaction_reserved_tokens": 40000
              },
              "timing_seconds": {
                "environment_setup": 28.872112,
                "agent_setup": 0.285374,
                "agent_execution": 1561.236596,
                "verifier": 25.458617
              },
              "usage": {
                "input_tokens": 6675352,
                "cached_tokens": 6621227,
                "output_tokens": 82701,
                "peak_context_tokens": 127610,
                "steps": 75,
                "requests": 76
              },
              "patch": {
                "bytes": 43693,
                "sha256": "70fb8c0bc110e5a688386a2a0215068e0493c584d16767026d32dc37538381b7"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [
            {
              "model": "gpt-6-astra",
              "provider": "openai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_gpt_6_astra_xhigh",
              "reasoning_effort": "xhigh",
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/awilix-async-container-initialization",
              "passed": 2,
              "attempts": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 2,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 3,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 4,
                  "outcome": "fail",
                  "passed": false
                }
              ]
            },
            {
              "model": "claude-opus-5",
              "provider": "vertex_ai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_claude_opus_5_max",
              "reasoning_effort": "max",
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/awilix-async-container-initialization",
              "passed": 0,
              "attempts": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 2,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 3,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 4,
                  "outcome": "fail",
                  "passed": false
                }
              ]
            }
          ],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "csstree-shorthand-expansion-compression",
          "official_prompt": {
            "title": "Add shorthand expansion and compression to the lexer",
            "description": "Add lexer methods to expand CSS shorthands into longhands and compress longhands back into shorthand values.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/csstree-shorthand-expansion-compression",
            "instruction_sha256": "fcf3984dab65cdb6fa5255587e5f09f75231da6c8e4840f47e25a4c0303afdb7",
            "task_toml_sha256": "461fbe15ef2afa084f988e225fab1bfafeb02d7f34114b55e01fefbd66ee31e6",
            "language": "javascript",
            "category": "feature_request"
          },
          "repository": "https://github.com/csstree/csstree",
          "base_revision": "88e3d965c0b1628642a30a841745b410d6835052",
          "runs": [
            {
              "run_id": "tonight-20261013/csstree-shorthand-expansion-comp__SoYR22E",
              "task_id": "csstree-shorthand-expansion-compression",
              "campaign": "tonight-20261013",
              "outcome": "graded-model-failure",
              "score": {
                "feature": {
                  "passed": 73,
                  "total": 79
                },
                "preservation": {
                  "passed": 16715,
                  "total": 16715
                },
                "reward": 0
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "agent": "OpenCode/Pier",
                "agent_version": "1.18.31",
                "context_tokens": 196608,
                "agent_timeout_seconds": 2700.0,
                "compaction_reserved_tokens": 40000
              },
              "timing_seconds": {
                "environment_setup": 25.6912,
                "agent_setup": 0.285557,
                "agent_execution": 996.228312,
                "verifier": 18.059544
              },
              "usage": {
                "input_tokens": 2902512,
                "cached_tokens": 2858140,
                "output_tokens": 55672,
                "peak_context_tokens": 84757,
                "steps": 52,
                "requests": 53
              },
              "patch": {
                "bytes": 41482,
                "sha256": "ae1e3b562991d498930c1d1563b6711b5e1fd6c97525827638c1a353773f89ad"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [
            {
              "model": "gpt-6-astra",
              "provider": "openai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_gpt_6_astra_xhigh",
              "reasoning_effort": "xhigh",
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/csstree-shorthand-expansion-compression",
              "passed": 2,
              "attempts": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 2,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 3,
                  "outcome": "pass",
                  "passed": true
                },
                {
                  "run_attempt": 4,
                  "outcome": "fail",
                  "passed": false
                }
              ]
            },
            {
              "model": "claude-opus-5",
              "provider": "vertex_ai",
              "harness": "mini-swe-agent",
              "source": "deep-swe",
              "eval_scope": "full",
              "config": "mini_swe_agent_claude_opus_5_max",
              "reasoning_effort": "max",
              "source_url": "https://deepswe.datacurve.ai/data/v1.1/tasks/csstree-shorthand-expansion-compression",
              "passed": 0,
              "attempts": 4,
              "attempt_outcomes": [
                {
                  "run_attempt": 1,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 2,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 3,
                  "outcome": "fail",
                  "passed": false
                },
                {
                  "run_attempt": 4,
                  "outcome": "fail",
                  "passed": false
                }
              ]
            }
          ],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        }
      ],
      "benchmark_revision": "0b9fabbb63b9104d678fe965e1632f2dd9eaa2ea"
    },
    {
      "id": "deepswe-deepswe-20261026",
      "label": "DeepSWE \u00b7 October 26 Flash-Next",
      "adapter": "deepswe",
      "benchmark": "DeepSWE v1.1",
      "campaign": "deepswe-20261026",
      "native_score_semantics": "feature-to-patch and patch-to-patch verifier checks",
      "tasks": [
        {
          "task_id": "oxvg-structural-selector-preservation",
          "official_prompt": {
            "title": "Preserve structure needed by stylesheet selectors",
            "description": "Keep rewrite optimizations from breaking structure-dependent stylesheet selector matching.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/oxvg-structural-selector-preservation",
            "instruction_sha256": "8d2295f55760d269913b22c17b4c345ca41fd2d586fc6ba59c34fb97b4422db9",
            "task_toml_sha256": "06e000bb4c397b72fcbf523d95923b73be636e8eb398b60842ab4f0e2ce57d23",
            "language": "rust",
            "category": "enhancement"
          },
          "repository": "https://github.com/noahbald/oxvg",
          "base_revision": "1fd7fab851ecc975e008be0e3e279568ce4e2b51",
          "runs": [
            {
              "run_id": "deepswe-20261026/oxvg-structural-selector-preserv__84qFU9b",
              "task_id": "oxvg-structural-selector-preservation",
              "campaign": "deepswe-20261026",
              "outcome": "pass",
              "score": {
                "feature": {
                  "passed": 6,
                  "total": 6
                },
                "preservation": {
                  "passed": 62,
                  "total": 62
                },
                "reward": 1
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "context": 196608,
                "engine": "Strata 0.1.48.1 + HC_Q8 (engine-only upgrade; existing server retained)",
                "quantization": "IQ3_S",
                "kv_cache": "int8",
                "mtp": 4,
                "vision": "CPU helper",
                "compaction_reserved_tokens": 40000,
                "agent_timeout_seconds": 2700,
                "output_policy": "Native remaining-context fitting; captured max_tokens is total capacity, not a fixed output grant"
              },
              "timing_seconds": {
                "environment_setup": 32.17989,
                "agent_setup": 0.297487,
                "agent_execution": 2001.737893,
                "verifier": 183.311801
              },
              "usage": {
                "input_tokens": 15766833,
                "cached_tokens": 15548491,
                "output_tokens": 83263,
                "peak_context_tokens": 156669,
                "steps": 176,
                "requests": 177
              },
              "patch": {
                "bytes": 34996,
                "sha256": "a9af42c105f68048c4b45c4616c4632bab32d660e9fab023fceebd514d263c5a"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "arcane-drift-detection-baselines",
          "official_prompt": {
            "title": "Add drift detection and compliance baselines",
            "description": "Implement baseline capture, drift comparison, and compliance tracking for container configurations.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/arcane-drift-detection-baselines",
            "instruction_sha256": "ca4de99c747ee4897ecaef75bbfc1324e147bd701958f1beb2473c76521b70e2",
            "task_toml_sha256": "bda69894e0fd8a7afda5dc81169d11db83058e78099c7889088f0c423a1db0a2",
            "language": "go",
            "category": "feature_request"
          },
          "repository": "https://github.com/getarcaneapp/arcane.git",
          "base_revision": "d34a5e2a6c5eb0f0955039775f5b9538424b58ff",
          "runs": [
            {
              "run_id": "deepswe-20261026/arcane-drift-detection-baselines__b8U72SC",
              "task_id": "arcane-drift-detection-baselines",
              "campaign": "deepswe-20261026",
              "outcome": "graded-model-failure",
              "score": {
                "feature": {
                  "passed": 81,
                  "total": 82
                },
                "preservation": {
                  "passed": 2,
                  "total": 2
                },
                "reward": 0
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "context": 196608,
                "engine": "Strata 0.1.48.1 + HC_Q8 (engine-only upgrade; existing server retained)",
                "quantization": "IQ3_S",
                "kv_cache": "int8",
                "mtp": 4,
                "vision": "CPU helper",
                "compaction_reserved_tokens": 40000,
                "agent_timeout_seconds": 2700,
                "output_policy": "Native remaining-context fitting; captured max_tokens is total capacity, not a fixed output grant"
              },
              "timing_seconds": {
                "environment_setup": 53.223699,
                "agent_setup": 0.280404,
                "agent_execution": 1152.519899,
                "verifier": 148.887855
              },
              "usage": {
                "input_tokens": 6160230,
                "cached_tokens": 6095201,
                "output_tokens": 49890,
                "peak_context_tokens": 114275,
                "steps": 78,
                "requests": 79
              },
              "patch": {
                "bytes": 59432,
                "sha256": "261f1d0044e67c7001435e6beddf22efa69fe445f0087eebf3bfaa2003c23bdd"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "bandit-interprocedural-taint-checks",
          "official_prompt": {
            "title": "Add interprocedural taint checks for Bandit injection sinks",
            "description": "Track user input through variables and calls so Bandit flags injection sinks beyond string literals.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/bandit-interprocedural-taint-checks",
            "instruction_sha256": "d780c0ef58de282522a94650d7fd54eed4e24aec73affc55c90898e2f11ba3b2",
            "task_toml_sha256": "eb38b1482426d94b804589a67b1632027c0ed9767b41524aabbf2472a2aca067",
            "language": "python",
            "category": "feature_request"
          },
          "repository": "https://github.com/PyCQA/bandit.git",
          "base_revision": "b46fa3a2723635aa29cc012538df4867ac2ac006",
          "runs": [
            {
              "run_id": "deepswe-20261026/bandit-interprocedural-taint-che__CLYFaWG",
              "task_id": "bandit-interprocedural-taint-checks",
              "campaign": "deepswe-20261026",
              "outcome": "graded-model-failure",
              "score": {
                "feature": {
                  "passed": 64,
                  "total": 66
                },
                "preservation": {
                  "passed": 293,
                  "total": 293
                },
                "reward": 0
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "context": 196608,
                "engine": "Strata 0.1.48.1 + HC_Q8 (engine-only upgrade; existing server retained)",
                "quantization": "IQ3_S",
                "kv_cache": "int8",
                "mtp": 4,
                "vision": "CPU helper",
                "compaction_reserved_tokens": 40000,
                "agent_timeout_seconds": 2700,
                "output_policy": "Native remaining-context fitting; captured max_tokens is total capacity, not a fixed output grant"
              },
              "timing_seconds": {
                "environment_setup": 24.558111,
                "agent_setup": 0.286798,
                "agent_execution": 783.459894,
                "verifier": 27.283702
              },
              "usage": {
                "input_tokens": 3655545,
                "cached_tokens": 3622672,
                "output_tokens": 40188,
                "peak_context_tokens": 70975,
                "steps": 74,
                "requests": 75
              },
              "patch": {
                "bytes": 29196,
                "sha256": "addccf3f13bd8f666620bb9eed5052928dcbf86cf65251ac6a3a3a792db35cf5"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "katex-multicolumn-array-spans",
          "official_prompt": {
            "title": "Add `\\multicolumn` column spans to array-like environments",
            "description": "Add `\\multicolumn` parsing and rendering for array-like environments with span-aware alignment and errors.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/katex-multicolumn-array-spans",
            "instruction_sha256": "54e5a648f5c2151a57f895d5ea4d814f44cca52daa5e58a198877f398ace28f1",
            "task_toml_sha256": "cd0d93fd52c3755d8ab44b7709253aca2b31d6eedcf7a33e15f856e5a35975b6",
            "language": "javascript",
            "category": "feature_request"
          },
          "repository": "https://github.com/KaTeX/KaTeX",
          "base_revision": "89bede495dc2c85e1c57ba627a18526f71d57396",
          "runs": [
            {
              "run_id": "deepswe-20261026/katex-multicolumn-array-spans__gQ4afLG",
              "task_id": "katex-multicolumn-array-spans",
              "campaign": "deepswe-20261026",
              "outcome": "graded-model-failure",
              "score": {
                "feature": {
                  "passed": 90,
                  "total": 94
                },
                "preservation": {
                  "passed": 599,
                  "total": 599
                },
                "reward": 0
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "context": 196608,
                "engine": "Strata 0.1.48.1 + HC_Q8 (engine-only upgrade; existing server retained)",
                "quantization": "IQ3_S",
                "kv_cache": "int8",
                "mtp": 4,
                "vision": "CPU helper",
                "compaction_reserved_tokens": 40000,
                "agent_timeout_seconds": 2700,
                "output_policy": "Native remaining-context fitting; captured max_tokens is total capacity, not a fixed output grant"
              },
              "timing_seconds": {
                "environment_setup": 32.097129,
                "agent_setup": 0.289844,
                "agent_execution": 1918.675055,
                "verifier": 20.529374
              },
              "usage": {
                "input_tokens": 11741827,
                "cached_tokens": 11543964,
                "output_tokens": 96376,
                "peak_context_tokens": 156602,
                "steps": 125,
                "requests": 126
              },
              "patch": {
                "bytes": 43508,
                "sha256": "cbeae09b3d1873e7ce8c9b47b9c423f584c5762a259c903cdc4a143571d9c59f"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "dynamodb-toolbox-lazy-recursive-schemas",
          "official_prompt": {
            "title": "Add lazy recursive schemas with DTO and JSON Schema export",
            "description": "Add a lazy schema type for self-referencing recursive data with full serialization, validation, and export support.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/dynamodb-toolbox-lazy-recursive-schemas",
            "instruction_sha256": "531983ffe0ed3b3a6dbc84deb6004d6ece59daa54830c360649d4c52977b0c9c",
            "task_toml_sha256": "9d644d0a98f9a8a9a8ec13de600ecdf5914c3aaaded2d37f8d0ea0e4828d4857",
            "language": "typescript",
            "category": "feature_request"
          },
          "repository": "https://github.com/dynamodb-toolbox/dynamodb-toolbox",
          "base_revision": "1f2a18664f8aded292707fcafb01ff15ea33d3b8",
          "runs": [
            {
              "run_id": "deepswe-20261026/dynamodb-toolbox-lazy-recursive__4HetmkQ",
              "task_id": "dynamodb-toolbox-lazy-recursive-schemas",
              "campaign": "deepswe-20261026",
              "outcome": "graded-model-failure",
              "score": {
                "feature": {
                  "passed": 35,
                  "total": 37
                },
                "preservation": {
                  "passed": 1267,
                  "total": 1267
                },
                "reward": 0
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "context": 196608,
                "engine": "Strata 0.1.48.1 + HC_Q8 (engine-only upgrade; existing server retained)",
                "quantization": "IQ3_S",
                "kv_cache": "int8",
                "mtp": 4,
                "vision": "CPU helper",
                "compaction_reserved_tokens": 40000,
                "agent_timeout_seconds": 2700,
                "output_policy": "Native remaining-context fitting; captured max_tokens is total capacity, not a fixed output grant"
              },
              "timing_seconds": {
                "environment_setup": 4.002591,
                "agent_setup": 0.286959,
                "agent_execution": 2199.150655,
                "verifier": 42.889821
              },
              "usage": {
                "input_tokens": 13636684,
                "cached_tokens": 13439239,
                "output_tokens": 93850,
                "peak_context_tokens": 156779,
                "steps": 177,
                "requests": 178
              },
              "patch": {
                "bytes": 95569,
                "sha256": "a0bb38b6b375629017b03e4a4b973175ff818d8f860163cac7a5705ae1e37aa9"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "pebble-durability-wait-apis",
          "official_prompt": {
            "title": "Add durability callbacks and wait APIs for sync writes",
            "description": "Add batch durability callbacks, durability wait methods, notifications, and metrics for sync write commits.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/pebble-durability-wait-apis",
            "instruction_sha256": "c450d7ce94f6312c9a1171eed84e0b7ab9bdc1c4cf9d79e5995c208a50c1c9b8",
            "task_toml_sha256": "e378b9c84c87b65ab26b7f6687fdebb8124c404654f17ee181cdafffe62910ce",
            "language": "go",
            "category": "feature_request"
          },
          "repository": "https://github.com/cockroachdb/pebble",
          "base_revision": "1454d2bc0f378d7f34766afafee68a77e7b85995",
          "runs": [
            {
              "run_id": "deepswe-20261026/pebble-durability-wait-apis__YEWCPE6",
              "task_id": "pebble-durability-wait-apis",
              "campaign": "deepswe-20261026",
              "outcome": "graded-model-failure",
              "score": {
                "feature": {
                  "passed": 0,
                  "total": 59
                },
                "preservation": {
                  "passed": 44,
                  "total": 44
                },
                "reward": 0
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "context": 196608,
                "engine": "Strata 0.1.48.1 + HC_Q8 (engine-only upgrade; existing server retained)",
                "quantization": "IQ3_S",
                "kv_cache": "int8",
                "mtp": 4,
                "vision": "CPU helper",
                "compaction_reserved_tokens": 40000,
                "agent_timeout_seconds": 2700,
                "output_policy": "Native remaining-context fitting; captured max_tokens is total capacity, not a fixed output grant",
                "delivery_note": "Empty submitted patch: solver ended by asking whether to commit. Official zero grades the unchanged baseline, not its unsubmitted draft. Delivery failure, not a completed feature implementation."
              },
              "timing_seconds": {
                "environment_setup": 27.138434,
                "agent_setup": 0.287084,
                "agent_execution": 2563.73458,
                "verifier": 101.832227
              },
              "usage": {
                "input_tokens": 13145450,
                "cached_tokens": 12919027,
                "output_tokens": 95526,
                "peak_context_tokens": 154314,
                "steps": 151,
                "requests": 152
              },
              "patch": {
                "bytes": 0,
                "sha256": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        }
      ],
      "benchmark_revision": "0b9fabbb63b9104d678fe965e1632f2dd9eaa2ea"
    },
    {
      "id": "rails-tonight-20261026",
      "label": "Rails \u00b7 tonight-20261026",
      "adapter": "rails-lemans",
      "benchmark": "Rails AI Evals",
      "campaign": "tonight-20261026",
      "native_score_semantics": "official verifier checks passed / checks collected",
      "tasks": [
        {
          "task_id": "ft-active-storage-tracking",
          "official_prompt": {
            "title": "Track storage usage for accounts and boards.",
            "description": "We need a way to track storage usage for each account and board. We only need to count card images and files embedded in board descriptions, card descriptions, and comments. A board's usage should include everything counted on that board and its cards. An account's usage should include everything counted across all of its boards.",
            "description_source": "official task prompt",
            "url": "https://raw.githubusercontent.com/rails/ai-evals/b799f091cf138bbc880baa6423eccd0d26066dbb/tasks/ft-active-storage-tracking/instruction.md",
            "instruction_sha256": "0da6f1f26af2e0237dcfc23a1261664939bb05e396e2866211eaf36f90685f6f"
          },
          "application": "fizzy",
          "benchmark_revision": "b799f091cf138bbc880baa6423eccd0d26066dbb",
          "runs": [
            {
              "run_id": "rails-tonight-20261026/ft-active-storage-tracking",
              "task_id": "ft-active-storage-tracking",
              "campaign": "rails-tonight-20261026",
              "outcome": "graded-model-failure",
              "score": {
                "checks": {
                  "passed": 14,
                  "total": 15
                },
                "reward": 0
              },
              "configuration": {
                "max_tokens": 196608,
                "min_p": 0.0,
                "model": "openai/qwen3.8-flash-next-iq3_s#medium",
                "presence_penalty": 0.0,
                "reasoning_effort": "medium",
                "repetition_penalty": 1.0,
                "temperature": 1.0,
                "top_k": 20,
                "top_p": 0.95,
                "agent": "miniswen-installed",
                "lemans_version": "1.5.1",
                "context": 196608,
                "engine": "Strata 0.1.48.1 + HC_Q8 (engine-only upgrade; existing server retained)",
                "quantization": "IQ3_S",
                "kv_cache": "int8",
                "mtp": 4,
                "vision": "CPU helper",
                "output_policy": "Native remaining-context fitting; captured max_tokens is total capacity, not a fixed output grant"
              },
              "timing_seconds": {
                "agent_execution": 1948.064,
                "wall_clock": 1997.0
              },
              "usage": {
                "input_tokens": 10607020,
                "cached_tokens": 10537348,
                "output_tokens": 91388,
                "steps": 107
              },
              "patch": {
                "bytes": 26774,
                "sha256": "c1b5c10424e6abf81c04c53c431f26b26caea4cb020dcf61db6cdbaf548b1f73"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "Rails Lemans official verifier",
                "evidence": "verified campaign report and native verifier checks"
              }
            }
          ],
          "frontier": []
        },
        {
          "task_id": "ft-cancelled-account-cleanup",
          "official_prompt": {
            "title": "Permanently remove data belonging to cancelled accounts after a grace period.",
            "description": "Cancelling an account makes it inactive immediately, but right now its data is kept forever. Add a recurring `Account::IncinerateDueJob` that permanently removes all data associated with accounts canceled more than 30 days ago from the database.",
            "description_source": "official task prompt",
            "url": "https://raw.githubusercontent.com/rails/ai-evals/b799f091cf138bbc880baa6423eccd0d26066dbb/tasks/ft-cancelled-account-cleanup/instruction.md",
            "instruction_sha256": "0a390b2506c1dde4a4256ae535f6c5457655b0151e4578f5d0f13942edb71ffa"
          },
          "application": "fizzy",
          "benchmark_revision": "b799f091cf138bbc880baa6423eccd0d26066dbb",
          "runs": [
            {
              "run_id": "rails-tonight-20261026/ft-cancelled-account-cleanup",
              "task_id": "ft-cancelled-account-cleanup",
              "campaign": "rails-tonight-20261026",
              "outcome": "graded-model-failure",
              "score": {
                "checks": {
                  "passed": 2,
                  "total": 5
                },
                "reward": 0
              },
              "configuration": {
                "max_tokens": 196608,
                "min_p": 0.0,
                "model": "openai/qwen3.8-flash-next-iq3_s#medium",
                "presence_penalty": 0.0,
                "reasoning_effort": "medium",
                "repetition_penalty": 1.0,
                "temperature": 1.0,
                "top_k": 20,
                "top_p": 0.95,
                "agent": "miniswen-installed",
                "lemans_version": "1.5.1",
                "context": 196608,
                "engine": "Strata 0.1.48.1 + HC_Q8 (engine-only upgrade; existing server retained)",
                "quantization": "IQ3_S",
                "kv_cache": "int8",
                "mtp": 4,
                "vision": "CPU helper",
                "output_policy": "Native remaining-context fitting; captured max_tokens is total capacity, not a fixed output grant"
              },
              "timing_seconds": {
                "agent_execution": 431.487,
                "wall_clock": 471.6
              },
              "usage": {
                "input_tokens": 1130637,
                "cached_tokens": 1095335,
                "output_tokens": 18182,
                "steps": 36
              },
              "patch": {
                "bytes": 4991,
                "sha256": "f9765e270cfeececb0dcf51e760ba30075833f8503a4ba4c220c0dbc46cc7c30"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "Rails Lemans official verifier",
                "evidence": "verified campaign report and native verifier checks"
              }
            }
          ],
          "frontier": []
        },
        {
          "task_id": "ft-card-drop-settled-landing",
          "official_prompt": {
            "title": "Drag cards around the board, settled the moment you let go.",
            "description": "We want people to change a card's column or status by dragging it around the board. Columns keep their own order \u2014 dragging moves a card between places, not within one.",
            "description_source": "official task prompt",
            "url": "https://raw.githubusercontent.com/rails/ai-evals/b799f091cf138bbc880baa6423eccd0d26066dbb/tasks/ft-card-drop-settled-landing/instruction.md",
            "instruction_sha256": "8a270019d64dcd2819b8d2473ceef76444e951e62cda72095f496f4b61328ee6"
          },
          "application": "fizzy",
          "benchmark_revision": "b799f091cf138bbc880baa6423eccd0d26066dbb",
          "runs": [
            {
              "run_id": "rails-tonight-20261026/ft-card-drop-settled-landing",
              "task_id": "ft-card-drop-settled-landing",
              "campaign": "rails-tonight-20261026",
              "outcome": "pass",
              "score": {
                "checks": {
                  "passed": 5,
                  "total": 5
                },
                "reward": 1
              },
              "configuration": {
                "max_tokens": 196608,
                "min_p": 0.0,
                "model": "openai/qwen3.8-flash-next-iq3_s#medium",
                "presence_penalty": 0.0,
                "reasoning_effort": "medium",
                "repetition_penalty": 1.0,
                "temperature": 1.0,
                "top_k": 20,
                "top_p": 0.95,
                "agent": "miniswen-installed",
                "lemans_version": "1.5.1",
                "context": 196608,
                "engine": "Strata 0.1.48.1 + HC_Q8 (engine-only upgrade; existing server retained)",
                "quantization": "IQ3_S",
                "kv_cache": "int8",
                "mtp": 4,
                "vision": "CPU helper",
                "output_policy": "Native remaining-context fitting; captured max_tokens is total capacity, not a fixed output grant"
              },
              "timing_seconds": {
                "agent_execution": 1390.381,
                "wall_clock": 1447.2
              },
              "usage": {
                "input_tokens": 5493847,
                "cached_tokens": 5440622,
                "output_tokens": 62583,
                "steps": 85
              },
              "patch": {
                "bytes": 11572,
                "sha256": "8405783d8ccbfa2e83ffaec013028399101a8ab1824cea5da79c018a984d7b72"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "Rails Lemans official verifier",
                "evidence": "verified campaign report and native verifier checks"
              }
            }
          ],
          "frontier": []
        },
        {
          "task_id": "ft-card-stalled-quiet-clock",
          "official_prompt": {
            "title": "Badge cards that caught fire and then went quiet, and let people list them.",
            "description": "Some cards get a burst of attention and then go cold. We want to flag those.",
            "description_source": "official task prompt",
            "url": "https://raw.githubusercontent.com/rails/ai-evals/b799f091cf138bbc880baa6423eccd0d26066dbb/tasks/ft-card-stalled-quiet-clock/instruction.md",
            "instruction_sha256": "acb37251ee296a9c990819ffda4d97605b742d71b3ad35100345b6093ca0850e"
          },
          "application": "fizzy",
          "benchmark_revision": "b799f091cf138bbc880baa6423eccd0d26066dbb",
          "runs": [
            {
              "run_id": "rails-tonight-20261026/ft-card-stalled-quiet-clock",
              "task_id": "ft-card-stalled-quiet-clock",
              "campaign": "rails-tonight-20261026",
              "outcome": "graded-model-failure",
              "score": {
                "checks": {
                  "passed": 6,
                  "total": 7
                },
                "reward": 0
              },
              "configuration": {
                "max_tokens": 196608,
                "min_p": 0.0,
                "model": "openai/qwen3.8-flash-next-iq3_s#medium",
                "presence_penalty": 0.0,
                "reasoning_effort": "medium",
                "repetition_penalty": 1.0,
                "temperature": 1.0,
                "top_k": 20,
                "top_p": 0.95,
                "agent": "miniswen-installed",
                "lemans_version": "1.5.1",
                "context": 196608,
                "engine": "Strata 0.1.48.1 + HC_Q8 (engine-only upgrade; existing server retained)",
                "quantization": "IQ3_S",
                "kv_cache": "int8",
                "mtp": 4,
                "vision": "CPU helper",
                "output_policy": "Native remaining-context fitting; captured max_tokens is total capacity, not a fixed output grant"
              },
              "timing_seconds": {
                "agent_execution": 943.548,
                "wall_clock": 987.6
              },
              "usage": {
                "input_tokens": 3195600,
                "cached_tokens": 3150303,
                "output_tokens": 42158,
                "steps": 64
              },
              "patch": {
                "bytes": 15437,
                "sha256": "3ef442fe5fbc691856ea66539da473d94ea07896ae19785db5635eeecd5b3a43"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "Rails Lemans official verifier",
                "evidence": "verified campaign report and native verifier checks"
              }
            }
          ],
          "frontier": []
        },
        {
          "task_id": "ft-comment-reaction-live-echo",
          "official_prompt": {
            "title": "Let people react to comments, live for everyone, settled instantly for the reactor.",
            "description": "People want to react to comments \u2014 emoji or a short bit of text, 16 characters max.",
            "description_source": "official task prompt",
            "url": "https://raw.githubusercontent.com/rails/ai-evals/b799f091cf138bbc880baa6423eccd0d26066dbb/tasks/ft-comment-reaction-live-echo/instruction.md",
            "instruction_sha256": "93f985d7a3e3cbe2a5253c3f5bb041c9d79b5f3499964aa33b61e996d3490897"
          },
          "application": "fizzy",
          "benchmark_revision": "b799f091cf138bbc880baa6423eccd0d26066dbb",
          "runs": [
            {
              "run_id": "rails-tonight-20261026/ft-comment-reaction-live-echo",
              "task_id": "ft-comment-reaction-live-echo",
              "campaign": "rails-tonight-20261026",
              "outcome": "graded-model-failure",
              "score": {
                "checks": {
                  "passed": 4,
                  "total": 6
                },
                "reward": 0
              },
              "configuration": {
                "max_tokens": 196608,
                "min_p": 0.0,
                "model": "openai/qwen3.8-flash-next-iq3_s#medium",
                "presence_penalty": 0.0,
                "reasoning_effort": "medium",
                "repetition_penalty": 1.0,
                "temperature": 1.0,
                "top_k": 20,
                "top_p": 0.95,
                "agent": "miniswen-installed",
                "lemans_version": "1.5.1",
                "context": 196608,
                "engine": "Strata 0.1.48.1 + HC_Q8 (engine-only upgrade; existing server retained)",
                "quantization": "IQ3_S",
                "kv_cache": "int8",
                "mtp": 4,
                "vision": "CPU helper",
                "output_policy": "Native remaining-context fitting; captured max_tokens is total capacity, not a fixed output grant"
              },
              "timing_seconds": {
                "agent_execution": 1712.013,
                "wall_clock": 1794.9
              },
              "usage": {
                "input_tokens": 7766786,
                "cached_tokens": 7707995,
                "output_tokens": 70701,
                "steps": 106
              },
              "patch": {
                "bytes": 23087,
                "sha256": "4f38dd88ffa19e82cd677f61aae577a0b5af8031f8571ab74e16e5c9997df3d8"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "Rails Lemans official verifier",
                "evidence": "verified campaign report and native verifier checks"
              }
            }
          ],
          "frontier": []
        },
        {
          "task_id": "ft-email-code-sign-in",
          "official_prompt": {
            "title": "Sign in, sign up and join with a six-character code sent by email.",
            "description": "Add sign-in by email code. Today the only way in is a passkey, and the email field on the sign-in, sign-up and join pages just says it isn't available yet.",
            "description_source": "official task prompt",
            "url": "https://raw.githubusercontent.com/rails/ai-evals/b799f091cf138bbc880baa6423eccd0d26066dbb/tasks/ft-email-code-sign-in/instruction.md",
            "instruction_sha256": "4597c38614d2ceccd237408050a86c322fbe3ebe9f6be51eec398efdb66212a7"
          },
          "application": "fizzy",
          "benchmark_revision": "b799f091cf138bbc880baa6423eccd0d26066dbb",
          "runs": [
            {
              "run_id": "rails-tonight-20261026/ft-email-code-sign-in",
              "task_id": "ft-email-code-sign-in",
              "campaign": "rails-tonight-20261026",
              "outcome": "pass",
              "score": {
                "checks": {
                  "passed": 7,
                  "total": 7
                },
                "reward": 1
              },
              "configuration": {
                "max_tokens": 196608,
                "min_p": 0.0,
                "model": "openai/qwen3.8-flash-next-iq3_s#medium",
                "presence_penalty": 0.0,
                "reasoning_effort": "medium",
                "repetition_penalty": 1.0,
                "temperature": 1.0,
                "top_k": 20,
                "top_p": 0.95,
                "agent": "miniswen-installed",
                "lemans_version": "1.5.1",
                "context": 196608,
                "engine": "Strata 0.1.48.1 + HC_Q8 (engine-only upgrade; existing server retained)",
                "quantization": "IQ3_S",
                "kv_cache": "int8",
                "mtp": 4,
                "vision": "CPU helper",
                "output_policy": "Native remaining-context fitting; captured max_tokens is total capacity, not a fixed output grant"
              },
              "timing_seconds": {
                "agent_execution": 1868.356,
                "wall_clock": 1950.8
              },
              "usage": {
                "input_tokens": 9238372,
                "cached_tokens": 9175432,
                "output_tokens": 87128,
                "steps": 103
              },
              "patch": {
                "bytes": 27103,
                "sha256": "778adab583fde3b17ff4db66831f8bffa77292302e0604d27fb81d24fdef1367"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "Rails Lemans official verifier",
                "evidence": "verified campaign report and native verifier checks"
              }
            }
          ],
          "frontier": []
        },
        {
          "task_id": "ft-entropy-sweep",
          "official_prompt": {
            "title": "Large-account performance incident: the hourly auto-postpone sweep no longer fits its hour.",
            "description": "Our install runs the SQLite build, and it has grown past a hundred thousand cards. Last week the hourly entropy sweep stopped fitting in its hour: one run is still going when the next one starts. APM sample from this morning:",
            "description_source": "official task prompt",
            "url": "https://raw.githubusercontent.com/rails/ai-evals/b799f091cf138bbc880baa6423eccd0d26066dbb/tasks/ft-entropy-sweep/instruction.md",
            "instruction_sha256": "fde969b1a4a74699e06b133c8400a6904c0f5d171bff780234d32f6faf0b838f"
          },
          "application": "fizzy",
          "benchmark_revision": "b799f091cf138bbc880baa6423eccd0d26066dbb",
          "runs": [
            {
              "run_id": "rails-tonight-20261026/ft-entropy-sweep",
              "task_id": "ft-entropy-sweep",
              "campaign": "rails-tonight-20261026",
              "outcome": "graded-model-failure",
              "score": {
                "checks": {
                  "passed": 7,
                  "total": 8
                },
                "reward": 0
              },
              "configuration": {
                "max_tokens": 196608,
                "min_p": 0.0,
                "model": "openai/qwen3.8-flash-next-iq3_s#medium",
                "presence_penalty": 0.0,
                "reasoning_effort": "medium",
                "repetition_penalty": 1.0,
                "temperature": 1.0,
                "top_k": 20,
                "top_p": 0.95,
                "agent": "miniswen-installed",
                "lemans_version": "1.5.1",
                "context": 196608,
                "engine": "Strata 0.1.48.1 + HC_Q8 (engine-only upgrade; existing server retained)",
                "quantization": "IQ3_S",
                "kv_cache": "int8",
                "mtp": 4,
                "vision": "CPU helper",
                "output_policy": "Native remaining-context fitting; captured max_tokens is total capacity, not a fixed output grant"
              },
              "timing_seconds": {
                "agent_execution": 1033.286,
                "wall_clock": 1090.6
              },
              "usage": {
                "input_tokens": 2048997,
                "cached_tokens": 2014029,
                "output_tokens": 40811,
                "steps": 50
              },
              "patch": {
                "bytes": 2930,
                "sha256": "60280e624e59a37b829a55e960ece9def1ddd67c9d0e1d703ab29d08a132204f"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "Rails Lemans official verifier",
                "evidence": "verified campaign report and native verifier checks"
              }
            }
          ],
          "frontier": []
        },
        {
          "task_id": "ft-local-time-human-clock",
          "official_prompt": {
            "title": "Render times in the browser, in the viewer's timezone, the way a person says them.",
            "description": "Times in the app are rendered by the server, in the server's timezone. That's wrong for anyone anywhere else, and it kills page caching: everyone's clock ends up baked into HTML we'd like to share between people.",
            "description_source": "official task prompt",
            "url": "https://raw.githubusercontent.com/rails/ai-evals/b799f091cf138bbc880baa6423eccd0d26066dbb/tasks/ft-local-time-human-clock/instruction.md",
            "instruction_sha256": "ca67aba184d460f8c349b573a3396908502b384eed8e738e6c21763d1e2339ca"
          },
          "application": "fizzy",
          "benchmark_revision": "b799f091cf138bbc880baa6423eccd0d26066dbb",
          "runs": [
            {
              "run_id": "rails-tonight-20261026/ft-local-time-human-clock",
              "task_id": "ft-local-time-human-clock",
              "campaign": "rails-tonight-20261026",
              "outcome": "pass",
              "score": {
                "checks": {
                  "passed": 6,
                  "total": 6
                },
                "reward": 1
              },
              "configuration": {
                "max_tokens": 196608,
                "min_p": 0.0,
                "model": "openai/qwen3.8-flash-next-iq3_s#medium",
                "presence_penalty": 0.0,
                "reasoning_effort": "medium",
                "repetition_penalty": 1.0,
                "temperature": 1.0,
                "top_k": 20,
                "top_p": 0.95,
                "agent": "miniswen-installed",
                "lemans_version": "1.5.1",
                "context": 196608,
                "engine": "Strata 0.1.48.1 + HC_Q8 (engine-only upgrade; existing server retained)",
                "quantization": "IQ3_S",
                "kv_cache": "int8",
                "mtp": 4,
                "vision": "CPU helper",
                "output_policy": "Native remaining-context fitting; captured max_tokens is total capacity, not a fixed output grant"
              },
              "timing_seconds": {
                "agent_execution": 978.807,
                "wall_clock": 1032.3
              },
              "usage": {
                "input_tokens": 2348938,
                "cached_tokens": 2306003,
                "output_tokens": 41333,
                "steps": 48
              },
              "patch": {
                "bytes": 16710,
                "sha256": "4b82047f940aa82fe78722df288fd1267bd1b7c5eddf06c9a7a1374a4c59318e"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "Rails Lemans official verifier",
                "evidence": "verified campaign report and native verifier checks"
              }
            }
          ],
          "frontier": []
        },
        {
          "task_id": "ft-mysql-fulltext-search-foundation",
          "official_prompt": {
            "title": "Search cards, scoped to the boards a person currently has access to.",
            "description": "Right now there's no way to find a card by what's in it. Add a search page at `/search?q=term` that finds cards by title, description, or the body of any comment on them, newest first, with matched words wrapped in `<mark>`. People only ever see results from boards they currently have access to. Draft cards and their comments never turn up.",
            "description_source": "official task prompt",
            "url": "https://raw.githubusercontent.com/rails/ai-evals/b799f091cf138bbc880baa6423eccd0d26066dbb/tasks/ft-mysql-fulltext-search-foundation/instruction.md",
            "instruction_sha256": "7ccab7a65c314759323c8f791deb8eeca6f30b55aff47ca8a11e82d11e632a38"
          },
          "application": "fizzy",
          "benchmark_revision": "b799f091cf138bbc880baa6423eccd0d26066dbb",
          "runs": [
            {
              "run_id": "rails-tonight-20261026/ft-mysql-fulltext-search-foundation",
              "task_id": "ft-mysql-fulltext-search-foundation",
              "campaign": "rails-tonight-20261026",
              "outcome": "graded-model-failure",
              "score": {
                "checks": {
                  "passed": 10,
                  "total": 13
                },
                "reward": 0
              },
              "configuration": {
                "max_tokens": 196608,
                "min_p": 0.0,
                "model": "openai/qwen3.8-flash-next-iq3_s#medium",
                "presence_penalty": 0.0,
                "reasoning_effort": "medium",
                "repetition_penalty": 1.0,
                "temperature": 1.0,
                "top_k": 20,
                "top_p": 0.95,
                "agent": "miniswen-installed",
                "lemans_version": "1.5.1",
                "context": 196608,
                "engine": "Strata 0.1.48.1 + HC_Q8 (engine-only upgrade; existing server retained)",
                "quantization": "IQ3_S",
                "kv_cache": "int8",
                "mtp": 4,
                "vision": "CPU helper",
                "output_policy": "Native remaining-context fitting; captured max_tokens is total capacity, not a fixed output grant"
              },
              "timing_seconds": {
                "agent_execution": 788.882,
                "wall_clock": 830.3
              },
              "usage": {
                "input_tokens": 2508204,
                "cached_tokens": 2466359,
                "output_tokens": 36209,
                "steps": 52
              },
              "patch": {
                "bytes": 11039,
                "sha256": "ac7b5001cffbe2a3a570dbb5ae21d277ed548c562239515299d13773f0f00e2d"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "Rails Lemans official verifier",
                "evidence": "verified campaign report and native verifier checks"
              }
            }
          ],
          "frontier": []
        }
      ]
    },
    {
      "id": "rails-single-20261027",
      "label": "Rails \u00b7 single-20261027",
      "adapter": "rails-lemans",
      "benchmark": "Rails AI Evals",
      "campaign": "single-20261027",
      "native_score_semantics": "official verifier checks passed / checks collected",
      "tasks": [
        {
          "task_id": "sup-cache-library-digest",
          "official_prompt": {
            "title": "An unscoped cache key serves one user's library digest to everyone.",
            "description": "Two support tickets this week describe the same thing. Somebody opens their library overview and finds books they were never given access to sitting in it \u2014 both times just after a colleague had opened theirs. Neither could reproduce it an hour later, and nobody here can reproduce it at all.",
            "description_source": "official task prompt",
            "url": "https://raw.githubusercontent.com/rails/ai-evals/b799f091cf138bbc880baa6423eccd0d26066dbb/tasks/sup-cache-library-digest/instruction.md",
            "instruction_sha256": "bc7bbe9970c23c39f890013ddba9afbf8ab1e545687c2900a91de2e90d9124f7"
          },
          "application": "default",
          "benchmark_revision": "b799f091cf138bbc880baa6423eccd0d26066dbb",
          "runs": [
            {
              "run_id": "rails-single-20261027/sup-cache-library-digest",
              "task_id": "sup-cache-library-digest",
              "campaign": "rails-single-20261027",
              "outcome": "pass",
              "score": {
                "checks": {
                  "passed": 4,
                  "total": 4
                },
                "reward": 1
              },
              "configuration": {
                "max_tokens": 196608,
                "min_p": 0.0,
                "model": "openai/qwen3.8-flash-next-iq3_s#medium",
                "presence_penalty": 0.0,
                "reasoning_effort": "medium",
                "repetition_penalty": 1.0,
                "temperature": 1.0,
                "top_k": 20,
                "top_p": 0.95,
                "agent": "miniswen-installed",
                "lemans_version": "1.5.1",
                "context": 196608,
                "engine": "Strata 0.1.48.1 + HC_Q8 (engine-only upgrade; existing server retained)",
                "quantization": "IQ3_S",
                "kv_cache": "int8",
                "mtp": 4,
                "vision": "CPU helper",
                "output_policy": "Native remaining-context fitting; captured max_tokens is total capacity, not a fixed output grant"
              },
              "timing_seconds": {
                "agent_execution": 166.197,
                "wall_clock": 192.1
              },
              "usage": {
                "input_tokens": 85976,
                "cached_tokens": 79437,
                "output_tokens": 6839,
                "steps": 15
              },
              "patch": {
                "bytes": 2327,
                "sha256": "66a4d9b6e2174af76f9112d016e61229b10d874b94c1f7153580fcafd0d5a9c2"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "Rails Lemans official verifier",
                "evidence": "verified campaign report and native verifier checks"
              }
            }
          ],
          "frontier": []
        }
      ]
    },
    {
      "id": "rails-lemans-tonight-20261027",
      "label": "Rails Lemans \u00b7 October 27 overnight",
      "adapter": "rails-lemans",
      "benchmark": "Rails AI Evals",
      "campaign": "tonight-20261027",
      "native_score_semantics": "official verifier checks passed / checks collected",
      "tasks": [
        {
          "task_id": "ft-saas-billing",
          "official_prompt": {
            "title": "Bring the paid product back to hosted Fizzy - Stripe-driven subscriptions and plans, free-tier card and storage allowances enforced in the app, the upgrade path in settings, staff comps and overrides, and a signature-guarded webhook endpoint.",
            "description": "Hosted Fizzy is going paid. Add billing.",
            "description_source": "official task prompt",
            "url": "https://raw.githubusercontent.com/rails/ai-evals/b799f091cf138bbc880baa6423eccd0d26066dbb/tasks/ft-saas-billing/instruction.md",
            "instruction_sha256": "c9d0a100e188e22c5990c4c1168a979019c14fe3b76d5bdcd5a1b4f52e5a922f"
          },
          "application": "Fizzy",
          "benchmark_revision": "b799f091cf138bbc880baa6423eccd0d26066dbb",
          "runs": [
            {
              "run_id": "rails-lemans-tonight-20261027/ft-saas-billing",
              "task_id": "ft-saas-billing",
              "campaign": "rails-lemans-tonight-20261027",
              "outcome": "graded-model-failure",
              "score": {
                "checks": {
                  "passed": 7,
                  "total": 8
                },
                "reward": 0
              },
              "configuration": {
                "max_tokens": 196608,
                "min_p": 0.0,
                "model": "openai/qwen3.8-flash-next-iq3_s#medium",
                "presence_penalty": 0.0,
                "reasoning_effort": "medium",
                "repetition_penalty": 1.0,
                "temperature": 1.0,
                "top_k": 20,
                "top_p": 0.95,
                "agent": "miniswen-installed",
                "lemans_version": "1.5.1",
                "context": 196608,
                "engine": "Strata 0.1.48.1 HC_Q8",
                "output_policy": "native remaining context",
                "agent_seconds_limit": 5400
              },
              "timing_seconds": {
                "agent_execution": 2227.02,
                "wall_clock": 2304.2
              },
              "usage": {
                "input_tokens": 14464015,
                "cached_tokens": 14395373,
                "output_tokens": 96571,
                "steps": 144
              },
              "patch": {
                "bytes": 54578,
                "sha256": "5e82bac72ef1f51145a8276f9efe40326d4a1d23ae8ec63d0848bec495da265c"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "Rails Lemans official verifier",
                "evidence": "verified campaign report and native verifier checks"
              }
            }
          ],
          "frontier": []
        },
        {
          "task_id": "ft-web-push-device-delivery",
          "official_prompt": {
            "title": "Push notifications for the web app, readable on every device, once, never inward.",
            "description": "Add push notifications to the web app: whatever lands in someone's tray also reaches their phone and desktop. People turn it on for the device they're using from notification settings. The browser hands the app the subscription it got from the push service as `push_subscription[endpoint, p256dh_key, auth_key]`, posted to `/users/:user_id/push_subscriptions`. A push says what the tray says and opens the card it's about. Ops will provide the key pair as `VAPID_PUBLIC_KEY` and `VAPID_PRIVATE_KEY`.",
            "description_source": "official task prompt",
            "url": "https://raw.githubusercontent.com/rails/ai-evals/b799f091cf138bbc880baa6423eccd0d26066dbb/tasks/ft-web-push-device-delivery/instruction.md",
            "instruction_sha256": "3211cf2eb79690afc7839521bd882e3ff843c2192b77717de1b31d7c3cabe408"
          },
          "application": "Fizzy",
          "benchmark_revision": "b799f091cf138bbc880baa6423eccd0d26066dbb",
          "runs": [
            {
              "run_id": "rails-lemans-tonight-20261027/ft-web-push-device-delivery",
              "task_id": "ft-web-push-device-delivery",
              "campaign": "rails-lemans-tonight-20261027",
              "outcome": "graded-model-failure",
              "score": {
                "checks": {
                  "passed": 3,
                  "total": 7
                },
                "reward": 0
              },
              "configuration": {
                "max_tokens": 196608,
                "min_p": 0.0,
                "model": "openai/qwen3.8-flash-next-iq3_s#medium",
                "presence_penalty": 0.0,
                "reasoning_effort": "medium",
                "repetition_penalty": 1.0,
                "temperature": 1.0,
                "top_k": 20,
                "top_p": 0.95,
                "agent": "miniswen-installed",
                "lemans_version": "1.5.1",
                "context": 196608,
                "engine": "Strata 0.1.48.1 HC_Q8",
                "output_policy": "native remaining context",
                "agent_seconds_limit": 5400
              },
              "timing_seconds": {
                "agent_execution": 1213.567,
                "wall_clock": 1271.1
              },
              "usage": {
                "input_tokens": 6554785,
                "cached_tokens": 6502799,
                "output_tokens": 53911,
                "steps": 103
              },
              "patch": {
                "bytes": 31359,
                "sha256": "56a98900bd10f25e9f725b9a800259a1d289a63be7868b74d5d466a9870369d8"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "Rails Lemans official verifier",
                "evidence": "verified campaign report and native verifier checks"
              }
            }
          ],
          "frontier": []
        },
        {
          "task_id": "sec-audit-sweep",
          "official_prompt": {
            "title": "Close five security defects from a checklist that states properties, never bugs.",
            "description": "An outside firm spent a week on this application. The write-up is embargoed until the fixes are out, so what we have is their finding list. They come back to retest against exactly what is written here.",
            "description_source": "official task prompt",
            "url": "https://raw.githubusercontent.com/rails/ai-evals/b799f091cf138bbc880baa6423eccd0d26066dbb/tasks/sec-audit-sweep/instruction.md",
            "instruction_sha256": "7de9b2711f5a37a3f2682b46218e32726ef3126bd5b6d6ae7f2cb65ee3511c91"
          },
          "application": "Writebook",
          "benchmark_revision": "b799f091cf138bbc880baa6423eccd0d26066dbb",
          "runs": [
            {
              "run_id": "rails-lemans-tonight-20261027/sec-audit-sweep",
              "task_id": "sec-audit-sweep",
              "campaign": "rails-lemans-tonight-20261027",
              "outcome": "pass",
              "score": {
                "checks": {
                  "passed": 5,
                  "total": 5
                },
                "reward": 1
              },
              "configuration": {
                "max_tokens": 196608,
                "min_p": 0.0,
                "model": "openai/qwen3.8-flash-next-iq3_s#medium",
                "presence_penalty": 0.0,
                "reasoning_effort": "medium",
                "repetition_penalty": 1.0,
                "temperature": 1.0,
                "top_k": 20,
                "top_p": 0.95,
                "agent": "miniswen-installed",
                "lemans_version": "1.5.1",
                "context": 196608,
                "engine": "Strata 0.1.48.1 HC_Q8",
                "output_policy": "native remaining context",
                "agent_seconds_limit": 5400
              },
              "timing_seconds": {
                "agent_execution": 754.005,
                "wall_clock": 782.3
              },
              "usage": {
                "input_tokens": 1985633,
                "cached_tokens": 1941594,
                "output_tokens": 28432,
                "steps": 48
              },
              "patch": {
                "bytes": 10871,
                "sha256": "f2e110c310e8a5685c740361a58737806f437fa870139239abf3572d46876747"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "Rails Lemans official verifier",
                "evidence": "verified campaign report and native verifier checks"
              }
            }
          ],
          "frontier": []
        },
        {
          "task_id": "sup-legacy-conversions",
          "official_prompt": {
            "title": "Sweep every removed Active Support conversion, including the no-argument aliases.",
            "description": "After last month's Rails upgrade, opening a book's shelf report gives a 500:",
            "description_source": "official task prompt",
            "url": "https://raw.githubusercontent.com/rails/ai-evals/b799f091cf138bbc880baa6423eccd0d26066dbb/tasks/sup-legacy-conversions/instruction.md",
            "instruction_sha256": "cc74f0fed1b5dd5c7272e5bba8b210ae28dc88fbd658614eeb4415297d3175a3"
          },
          "application": "Writebook",
          "benchmark_revision": "b799f091cf138bbc880baa6423eccd0d26066dbb",
          "runs": [
            {
              "run_id": "rails-lemans-tonight-20261027/sup-legacy-conversions",
              "task_id": "sup-legacy-conversions",
              "campaign": "rails-lemans-tonight-20261027",
              "outcome": "pass",
              "score": {
                "checks": {
                  "passed": 6,
                  "total": 6
                },
                "reward": 1
              },
              "configuration": {
                "max_tokens": 196608,
                "min_p": 0.0,
                "model": "openai/qwen3.8-flash-next-iq3_s#medium",
                "presence_penalty": 0.0,
                "reasoning_effort": "medium",
                "repetition_penalty": 1.0,
                "temperature": 1.0,
                "top_k": 20,
                "top_p": 0.95,
                "agent": "miniswen-installed",
                "lemans_version": "1.5.1",
                "context": 196608,
                "engine": "Strata 0.1.48.1 HC_Q8",
                "output_policy": "native remaining context",
                "agent_seconds_limit": 5400
              },
              "timing_seconds": {
                "agent_execution": 446.796,
                "wall_clock": 474.6
              },
              "usage": {
                "input_tokens": 824942,
                "cached_tokens": 802215,
                "output_tokens": 19730,
                "steps": 33
              },
              "patch": {
                "bytes": 4842,
                "sha256": "6b1177cb2dc46d78117ee4c1e583d849427c32c6ed9f2e0a969ce4eba4c9b341"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "Rails Lemans official verifier",
                "evidence": "verified campaign report and native verifier checks"
              }
            }
          ],
          "frontier": []
        },
        {
          "task_id": "sup-log-to-terminal",
          "official_prompt": {
            "title": "Add terminal output to a runner script without losing its log output.",
            "description": "`script/library_report.rb` is the last step of our import runbook. When an import finishes, the on-call runs",
            "description_source": "official task prompt",
            "url": "https://raw.githubusercontent.com/rails/ai-evals/b799f091cf138bbc880baa6423eccd0d26066dbb/tasks/sup-log-to-terminal/instruction.md",
            "instruction_sha256": "3852bc5ab676b1f28983adb39af9e9cd660303860f5e5ce4eb58554c6b041fa1"
          },
          "application": "Writebook",
          "benchmark_revision": "b799f091cf138bbc880baa6423eccd0d26066dbb",
          "runs": [
            {
              "run_id": "rails-lemans-tonight-20261027/sup-log-to-terminal",
              "task_id": "sup-log-to-terminal",
              "campaign": "rails-lemans-tonight-20261027",
              "outcome": "pass",
              "score": {
                "checks": {
                  "passed": 3,
                  "total": 3
                },
                "reward": 1
              },
              "configuration": {
                "max_tokens": 196608,
                "min_p": 0.0,
                "model": "openai/qwen3.8-flash-next-iq3_s#medium",
                "presence_penalty": 0.0,
                "reasoning_effort": "medium",
                "repetition_penalty": 1.0,
                "temperature": 1.0,
                "top_k": 20,
                "top_p": 0.95,
                "agent": "miniswen-installed",
                "lemans_version": "1.5.1",
                "context": 196608,
                "engine": "Strata 0.1.48.1 HC_Q8",
                "output_policy": "native remaining context",
                "agent_seconds_limit": 5400
              },
              "timing_seconds": {
                "agent_execution": 241.965,
                "wall_clock": 272.6
              },
              "usage": {
                "input_tokens": 363783,
                "cached_tokens": 346541,
                "output_tokens": 10253,
                "steps": 24
              },
              "patch": {
                "bytes": 891,
                "sha256": "8e4ebf66aa7bba785bfc587688ade64614de5b246aa09f5f8db92b92393bc988"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "Rails Lemans official verifier",
                "evidence": "verified campaign report and native verifier checks"
              }
            }
          ],
          "frontier": []
        },
        {
          "task_id": "tst-error-page-flake",
          "official_prompt": {
            "title": "An order-dependent suite flake from a helper leaking a process-wide toggle.",
            "description": "Our full test runs have gone flaky. Every file passes on its own, but `bin/rails test` fails more runs than not, and which tests fail changes every run: sometimes controller tests that expect a rendered 404 page error out with a raised ActiveRecord::RecordNotFound instead; other runs those pass and test/integration/book_absence_test.rb fails with \"nothing was raised\".",
            "description_source": "official task prompt",
            "url": "https://raw.githubusercontent.com/rails/ai-evals/b799f091cf138bbc880baa6423eccd0d26066dbb/tasks/tst-error-page-flake/instruction.md",
            "instruction_sha256": "1dbbdc67ff5b7d455f2073c9cd81b81b1786b645798af2269508d7bf47ae9a8c"
          },
          "application": "Writebook",
          "benchmark_revision": "b799f091cf138bbc880baa6423eccd0d26066dbb",
          "runs": [
            {
              "run_id": "rails-lemans-tonight-20261027/tst-error-page-flake",
              "task_id": "tst-error-page-flake",
              "campaign": "rails-lemans-tonight-20261027",
              "outcome": "pass",
              "score": {
                "checks": {
                  "passed": 6,
                  "total": 6
                },
                "reward": 1
              },
              "configuration": {
                "max_tokens": 196608,
                "min_p": 0.0,
                "model": "openai/qwen3.8-flash-next-iq3_s#medium",
                "presence_penalty": 0.0,
                "reasoning_effort": "medium",
                "repetition_penalty": 1.0,
                "temperature": 1.0,
                "top_k": 20,
                "top_p": 0.95,
                "agent": "miniswen-installed",
                "lemans_version": "1.5.1",
                "context": 196608,
                "engine": "Strata 0.1.48.1 HC_Q8",
                "output_policy": "native remaining context",
                "agent_seconds_limit": 5400
              },
              "timing_seconds": {
                "agent_execution": 693.859,
                "wall_clock": 720.0
              },
              "usage": {
                "input_tokens": 630721,
                "cached_tokens": 619785,
                "output_tokens": 15895,
                "steps": 36
              },
              "patch": {
                "bytes": 1235,
                "sha256": "26430caabed37db20ac582e695cb9275b94e85ff58d0e41687db5b33f749cc01"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "Rails Lemans official verifier",
                "evidence": "verified campaign report and native verifier checks"
              }
            }
          ],
          "frontier": []
        }
      ]
    },
    {
      "id": "deepswe-tonight-20261027",
      "label": "DeepSWE \u00b7 October 27 overnight",
      "adapter": "deepswe",
      "benchmark": "DeepSWE v1.1",
      "campaign": "tonight-20261027",
      "native_score_semantics": "feature-to-patch and patch-to-patch verifier checks",
      "tasks": [
        {
          "task_id": "superjson-error-stack-serialization",
          "official_prompt": {
            "title": "Add error stack serialization to SuperJSON",
            "description": "Add configurable serialization and restoration of error stacks, stack frames, causes, and sanitization in SuperJSON.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/superjson-error-stack-serialization",
            "instruction_sha256": "cb5e04f0f6c01180a47f1205b6d86126a8a7f9e7d5662e4f9477daa862826e30",
            "task_toml_sha256": "c808502b1f72060e23e3ac7885b6edd8fa49ad68f2e175f017d560222145cd33",
            "language": "typescript",
            "category": "feature_request"
          },
          "repository": "https://github.com/flightcontrolhq/superjson.git",
          "base_revision": "010c4bdb4b8758844fd44eacf38e42b22eba8aea",
          "runs": [
            {
              "run_id": "tonight-20261027/superjson-error-stack-serializat__cjLWoxd",
              "task_id": "superjson-error-stack-serialization",
              "campaign": "tonight-20261027",
              "outcome": "pass",
              "score": {
                "feature": {
                  "passed": 80,
                  "total": 80
                },
                "preservation": {
                  "passed": 116,
                  "total": 116
                },
                "reward": 1
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "context": 196608,
                "engine": "Strata 0.1.48.1 HC_Q8",
                "output_policy": "native remaining context",
                "agent_seconds_limit": 2700
              },
              "timing_seconds": {
                "environment_setup": 28.617699,
                "agent_setup": 0.288276,
                "agent_execution": 620.337573,
                "verifier": 19.079745
              },
              "usage": {
                "input_tokens": 1497056,
                "cached_tokens": 1466426,
                "output_tokens": 36648,
                "peak_context_tokens": 62485,
                "steps": 32,
                "requests": 33
              },
              "patch": {
                "bytes": 38606,
                "sha256": "6c4c4863d3e10fd5489af912caebbcb0d85d3b787d31b387f34c2ee3509f73e6"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "ts-pattern-match-each",
          "official_prompt": {
            "title": "Add `matchEach` to ts-pattern",
            "description": "Add a new `matchEach` matcher that evaluates all matching clauses and returns all results in order.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/ts-pattern-match-each",
            "instruction_sha256": "fb3807619644dcb10abd97166980d79d69148729df24f97c41d40bd1b879a65f",
            "task_toml_sha256": "3a18942908d8e2441ebb9fc13da6ad6d8ea10cf4ecc9258689fd0a1f22a7bed2",
            "language": "typescript",
            "category": "feature_request"
          },
          "repository": "https://github.com/gvergnaud/ts-pattern",
          "base_revision": "f66fc061fde4f764b113ededa09be63dae564159",
          "runs": [
            {
              "run_id": "tonight-20261027/ts-pattern-match-each__mipJUBf",
              "task_id": "ts-pattern-match-each",
              "campaign": "tonight-20261027",
              "outcome": "graded-model-failure",
              "score": {
                "feature": {
                  "passed": 0,
                  "total": 85
                },
                "preservation": {
                  "passed": 6,
                  "total": 6
                },
                "reward": 0
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "context": 196608,
                "engine": "Strata 0.1.48.1 HC_Q8",
                "output_policy": "native remaining context",
                "agent_seconds_limit": 2700
              },
              "timing_seconds": {
                "environment_setup": 26.840494,
                "agent_setup": 0.306052,
                "agent_execution": 787.889527,
                "verifier": 20.449883
              },
              "usage": {
                "input_tokens": 2163035,
                "cached_tokens": 2130535,
                "output_tokens": 41807,
                "peak_context_tokens": 70788,
                "steps": 46,
                "requests": 47
              },
              "patch": {
                "bytes": 36467,
                "sha256": "ef8098c495d788a5edfebd8b795dba903f4671bb2655520186675ea8267b1da4"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "kea-atomic-signal-selectors",
          "official_prompt": {
            "title": "Add atomic signal selectors to Kea",
            "description": "Introduce fine-grained atomic selector tracking with dependency health, circular detection, and unchanged Kea lifecycle behavior.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/kea-atomic-signal-selectors",
            "instruction_sha256": "a7bde3311b63eb381fc29740e504ccb259e60466625b0626fa440ac545527deb",
            "task_toml_sha256": "e348f26d469643cf8bfd535e9d9f86c8c20f7e7169aeda515223874a9dea271c",
            "language": "typescript",
            "category": "feature_request"
          },
          "repository": "https://github.com/keajs/kea.git",
          "base_revision": "6c7ebba57821989733a11d6f3888816658584d97",
          "runs": [
            {
              "run_id": "tonight-20261027/kea-atomic-signal-selectors__UVSHKkm",
              "task_id": "kea-atomic-signal-selectors",
              "campaign": "tonight-20261027",
              "outcome": "pass",
              "score": {
                "feature": {
                  "passed": 12,
                  "total": 12
                },
                "preservation": {
                  "passed": 139,
                  "total": 139
                },
                "reward": 1
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "context": 196608,
                "engine": "Strata 0.1.48.1 HC_Q8",
                "output_policy": "native remaining context",
                "agent_seconds_limit": 2700
              },
              "timing_seconds": {
                "environment_setup": 29.377158,
                "agent_setup": 0.300392,
                "agent_execution": 1547.746277,
                "verifier": 22.360163
              },
              "usage": {
                "input_tokens": 4389980,
                "cached_tokens": 4324340,
                "output_tokens": 59727,
                "peak_context_tokens": 109070,
                "steps": 55,
                "requests": 56
              },
              "patch": {
                "bytes": 40795,
                "sha256": "47805eccc44db9d63bdfbfbf829fbea32669f7825e6429d0e1c92a86297b262d"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "kgateway-consistent-hash-policy",
          "official_prompt": {
            "title": "Add consistent hash policy support to TrafficPolicy",
            "description": "Add spec.consistentHash to TrafficPolicy and merge it into route hash_policy generation.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/kgateway-consistent-hash-policy",
            "instruction_sha256": "6316ebc35a9c06f88ad2a8a72aa395dc0059927cec392b820106b27e9c20aa3f",
            "task_toml_sha256": "7994028e8d6a5de468af7a97b9a4ede8a0dc5a9280eabac18406a2735e85fd68",
            "language": "go",
            "category": "feature_request"
          },
          "repository": "https://github.com/kgateway-dev/kgateway",
          "base_revision": "7abc5278782e3280fec8292b39807ec1b537eaf4",
          "runs": [
            {
              "run_id": "tonight-20261027/kgateway-consistent-hash-policy__9xkjaoa",
              "task_id": "kgateway-consistent-hash-policy",
              "campaign": "tonight-20261027",
              "outcome": "graded-model-failure",
              "score": {
                "feature": {
                  "passed": 0,
                  "total": 2
                },
                "preservation": {
                  "passed": 214,
                  "total": 214
                },
                "reward": 0
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "context": 196608,
                "engine": "Strata 0.1.48.1 HC_Q8",
                "output_policy": "native remaining context",
                "agent_seconds_limit": 2700
              },
              "timing_seconds": {
                "environment_setup": 44.610372,
                "agent_setup": 0.294499,
                "agent_execution": 1957.460358,
                "verifier": 189.951678
              },
              "usage": {
                "input_tokens": 12467393,
                "cached_tokens": 12387437,
                "output_tokens": 61236,
                "peak_context_tokens": 138222,
                "steps": 132,
                "requests": 133
              },
              "patch": {
                "bytes": 90069,
                "sha256": "7758ab1246626147c30abe03d605c78e2d78fb9e0fc9a11f72eeae235d3cae68"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "pwntools-tube-multiplexing",
          "official_prompt": {
            "title": "Add tube multiplexing to pwntools",
            "description": "Add a TubeMultiplexer and MuxChannel API for multiple bidirectional logical channels over one tube, with flow control and close propagation.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/pwntools-tube-multiplexing",
            "instruction_sha256": "d57d68f683be7ddebfb65a2a25dd139143ebc4409c46f09425ea4e38ef3b336a",
            "task_toml_sha256": "745d238dd73edca759016a833da57c236901f1461416a2d46844c2bfa933a048",
            "language": "python",
            "category": "feature_request"
          },
          "repository": "https://github.com/Gallopsled/pwntools",
          "base_revision": "76894a5404a65d2800b6d0adaf3485ecba275caa",
          "runs": [
            {
              "run_id": "tonight-20261027/pwntools-tube-multiplexing__an5KMib",
              "task_id": "pwntools-tube-multiplexing",
              "campaign": "tonight-20261027",
              "outcome": "graded-model-failure",
              "score": {
                "feature": {
                  "passed": 67,
                  "total": 73
                },
                "preservation": {
                  "passed": 1,
                  "total": 1
                },
                "reward": 0
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "context": 196608,
                "engine": "Strata 0.1.48.1 HC_Q8",
                "output_policy": "native remaining context",
                "agent_seconds_limit": 2700
              },
              "timing_seconds": {
                "environment_setup": 27.227184,
                "agent_setup": 0.29218,
                "agent_execution": 1777.42547,
                "verifier": 45.267786
              },
              "usage": {
                "input_tokens": 8494872,
                "cached_tokens": 8363050,
                "output_tokens": 90626,
                "peak_context_tokens": 141579,
                "steps": 93,
                "requests": 94
              },
              "patch": {
                "bytes": 52833,
                "sha256": "feec963f9d82d537ab5c02a9f71a4700c99e0152c740a8873d956c748415d5c1"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "query-persist-restored-query-state",
          "official_prompt": {
            "title": "Preserve restored query state in persisted snapshots",
            "description": "Preserve full persisted query state, including errors, counters, timestamps, and infinite pagination, during restoration and cache rebuilds.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/query-persist-restored-query-state",
            "instruction_sha256": "f3a5f63f195bbc7d556ff62ffb609e2bef69cbef25f1e0c537b8d18cfa91ac95",
            "task_toml_sha256": "d3ffb5325266363d392750f96eb2ad9a453a8252ba6dac2d4cbd9538b16d9276",
            "language": "typescript",
            "category": "feature_request"
          },
          "repository": "https://github.com/TanStack/query",
          "base_revision": "1047cdc393fac7c98822c993d70c28f58833c63d",
          "runs": [
            {
              "run_id": "tonight-20261027/query-persist-restored-query-sta__6TKbZEk",
              "task_id": "query-persist-restored-query-state",
              "campaign": "tonight-20261027",
              "outcome": "pass",
              "score": {
                "feature": {
                  "passed": 8,
                  "total": 8
                },
                "preservation": {
                  "passed": 42,
                  "total": 42
                },
                "reward": 1
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "context": 196608,
                "engine": "Strata 0.1.48.1 HC_Q8",
                "output_policy": "native remaining context",
                "agent_seconds_limit": 2700
              },
              "timing_seconds": {
                "environment_setup": 70.520461,
                "agent_setup": 0.300628,
                "agent_execution": 1350.073237,
                "verifier": 24.363802
              },
              "usage": {
                "input_tokens": 8560354,
                "cached_tokens": 8487329,
                "output_tokens": 62791,
                "peak_context_tokens": 135817,
                "steps": 96,
                "requests": 97
              },
              "patch": {
                "bytes": 36659,
                "sha256": "84705b6f6aef134892251022ac4e988e31ca433857809b8e3923c31c179e1c73"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "textual-kitty-key-phases",
          "official_prompt": {
            "title": "Complete Kitty keyboard phases and stable fallback key metadata",
            "description": "Add full Kitty keyboard phase handling and stable fallback key metadata for keys and shortcuts.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/textual-kitty-key-phases",
            "instruction_sha256": "88a48b1028f82a815f135e122e9b57cc0e942c677cf0dfe49cbc93318e1c5093",
            "task_toml_sha256": "0470d205fd7feb2177ad8a2b1741c35c3ed179cdef9430c781f33aba141f98bd",
            "language": "python",
            "category": "feature_request"
          },
          "repository": "https://github.com/Textualize/textual",
          "base_revision": "9737a5ab723f79e59f0a83eb036a3d15fad6b054",
          "runs": [
            {
              "run_id": "tonight-20261027/textual-kitty-key-phases__G8ZBERf",
              "task_id": "textual-kitty-key-phases",
              "campaign": "tonight-20261027",
              "outcome": "graded-model-failure",
              "score": {
                "feature": {
                  "passed": 19,
                  "total": 23
                },
                "preservation": {
                  "passed": 57,
                  "total": 57
                },
                "reward": 0
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "context": 196608,
                "engine": "Strata 0.1.48.1 HC_Q8",
                "output_policy": "native remaining context",
                "agent_seconds_limit": 2700
              },
              "timing_seconds": {
                "environment_setup": 32.229703,
                "agent_setup": 0.29841,
                "agent_execution": 1840.347013,
                "verifier": 18.368347
              },
              "usage": {
                "input_tokens": 4706681,
                "cached_tokens": 4672645,
                "output_tokens": 43621,
                "peak_context_tokens": 83426,
                "steps": 91,
                "requests": 92
              },
              "patch": {
                "bytes": 19214,
                "sha256": "16bcf967fc8a5c6d1a3737939efe3e01367210c568cda41458e5a106a4ff3ce4"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "arktype-json-schema-refs-dependencies",
          "official_prompt": {
            "title": "Add JSON Schema refs and dependency keywords",
            "description": "Add JSON Schema dependency keywords, local $defs/$ref resolution, and conditional schema handling.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/arktype-json-schema-refs-dependencies",
            "instruction_sha256": "594fa84b95c92f69e949de2781107a3ff5b6131e889203fc4505f6f41a136ea3",
            "task_toml_sha256": "6a91e7bd0fd5d1154f77e84c69a62d266e24552c83954dec0f18d3471b6a809c",
            "language": "typescript",
            "category": "feature_request"
          },
          "repository": "https://github.com/arktypeio/arktype",
          "base_revision": "04355e8b26d1ad5264ef62314a2bc46c4de58ed8",
          "runs": [
            {
              "run_id": "tonight-20261027/arktype-json-schema-refs-depende__yF6x8bz",
              "task_id": "arktype-json-schema-refs-dependencies",
              "campaign": "tonight-20261027",
              "outcome": "pass",
              "score": {
                "feature": {
                  "passed": 25,
                  "total": 25
                },
                "preservation": {
                  "passed": 1679,
                  "total": 1679
                },
                "reward": 1
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "context": 196608,
                "engine": "Strata 0.1.48.1 HC_Q8",
                "output_policy": "native remaining context",
                "agent_seconds_limit": 2700
              },
              "timing_seconds": {
                "environment_setup": 46.294337,
                "agent_setup": 0.29379,
                "agent_execution": 1565.160388,
                "verifier": 59.424845
              },
              "usage": {
                "input_tokens": 10459775,
                "cached_tokens": 10404635,
                "output_tokens": 76131,
                "peak_context_tokens": 137803,
                "steps": 119,
                "requests": 120
              },
              "patch": {
                "bytes": 34842,
                "sha256": "63f5cb1845a848ee93e73f15f213a682be88843d1a79e62d3bbbaa21221bad1c"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "claude-code-by-agents-recursive-delegation",
          "official_prompt": {
            "title": "Implement recursive agent delegation through delegate_task tool calls",
            "description": "Run delegated sub-agents, feed their results back as tool_result, and handle unknown agents, failures, and circular delegation.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/claude-code-by-agents-recursive-delegation",
            "instruction_sha256": "56818b3491a387c482911fe0a88859cb1b802e56a7087d51e0a77965fd68a59c",
            "task_toml_sha256": "5f5b092341a9f654ac3b25c68381f6a0eb028280f6fbc4001003227e4729959d",
            "language": "typescript",
            "category": "feature_request"
          },
          "repository": "https://github.com/baryhuang/claude-code-by-agents",
          "base_revision": "5e0a2247d446c49a9951a06bb83b6e956dc7eb41",
          "runs": [
            {
              "run_id": "tonight-20261027/claude-code-by-agents-recursive__sfSLPcY",
              "task_id": "claude-code-by-agents-recursive-delegation",
              "campaign": "tonight-20261027",
              "outcome": "pass",
              "score": {
                "feature": {
                  "passed": 7,
                  "total": 7
                },
                "preservation": {
                  "passed": 31,
                  "total": 31
                },
                "reward": 1
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "context": 196608,
                "engine": "Strata 0.1.48.1 HC_Q8",
                "output_policy": "native remaining context",
                "agent_seconds_limit": 2700
              },
              "timing_seconds": {
                "environment_setup": 35.832673,
                "agent_setup": 0.286154,
                "agent_execution": 1000.997591,
                "verifier": 18.694793
              },
              "usage": {
                "input_tokens": 1949544,
                "cached_tokens": 1901483,
                "output_tokens": 41513,
                "peak_context_tokens": 89280,
                "steps": 30,
                "requests": 52
              },
              "patch": {
                "bytes": 36779,
                "sha256": "314cec6353e49d4745dd6d8bbbaea66ba78f5446b9b086bd4c32bd40c1d7c1a6"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "clack-async-autocomplete-options",
          "official_prompt": {
            "title": "Add async autocomplete options and fetch lifecycle handling",
            "description": "Add async option fetching with caching, retries, debouncing, and loading state to AutocompletePrompt.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/clack-async-autocomplete-options",
            "instruction_sha256": "5792630a916783910328e8d7778dc9483ed949bf713a09d00d4b8f6d4bdb6bf3",
            "task_toml_sha256": "977c453c2c4701b1d58df42f7d11a9f9b16ae9b57c7e2522fa45c2bd14f44603",
            "language": "typescript",
            "category": "feature_request"
          },
          "repository": "https://github.com/bombshell-dev/clack",
          "base_revision": "8a96e2dcd7f821d1250b58cf71c327679f94de25",
          "runs": [
            {
              "run_id": "tonight-20261027/clack-async-autocomplete-options__7r9i8KP",
              "task_id": "clack-async-autocomplete-options",
              "campaign": "tonight-20261027",
              "outcome": "pass",
              "score": {
                "feature": {
                  "passed": 82,
                  "total": 82
                },
                "preservation": {
                  "passed": 643,
                  "total": 643
                },
                "reward": 1
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "context": 196608,
                "engine": "Strata 0.1.48.1 HC_Q8",
                "output_policy": "native remaining context",
                "agent_seconds_limit": 2700
              },
              "timing_seconds": {
                "environment_setup": 31.66658,
                "agent_setup": 0.296273,
                "agent_execution": 1189.531599,
                "verifier": 30.700203
              },
              "usage": {
                "input_tokens": 5478704,
                "cached_tokens": 5441175,
                "output_tokens": 64300,
                "peak_context_tokens": 101363,
                "steps": 83,
                "requests": 84
              },
              "patch": {
                "bytes": 60154,
                "sha256": "f817037e65b71a316756da8f94ac0a2a6d0f49b23a2f5523b8b5f989775abb0d"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "scc-bounded-memory-spilling",
          "official_prompt": {
            "title": "Add bounded-memory spilling to SCC aggregation",
            "description": "Add an opt-in bounded-memory mode that spills per-file aggregation results to disk without changing output.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/scc-bounded-memory-spilling",
            "instruction_sha256": "fc99a21974fa57940e3ff7e9eb81c31b06728cbc39177faf5500667cd51a816e",
            "task_toml_sha256": "40df2d49f62c899ba5dd08a99ba9ee3de09cea150f6bb69aac76f48b5c0f33b5",
            "language": "go",
            "category": "feature_request"
          },
          "repository": "https://github.com/boyter/scc.git",
          "base_revision": "bc2796e01998ebc2d40818323f93113aed2542ea",
          "runs": [
            {
              "run_id": "tonight-20261027/scc-bounded-memory-spilling__Sh8hyEL",
              "task_id": "scc-bounded-memory-spilling",
              "campaign": "tonight-20261027",
              "outcome": "graded-model-failure",
              "score": {
                "feature": {
                  "passed": 28,
                  "total": 31
                },
                "preservation": {
                  "passed": 286,
                  "total": 286
                },
                "reward": 0
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "context": 196608,
                "engine": "Strata 0.1.48.1 HC_Q8",
                "output_policy": "native remaining context",
                "agent_seconds_limit": 2700
              },
              "timing_seconds": {
                "environment_setup": 28.715387,
                "agent_setup": 0.291816,
                "agent_execution": 750.692371,
                "verifier": 42.356515
              },
              "usage": {
                "input_tokens": 2755869,
                "cached_tokens": 2694261,
                "output_tokens": 39927,
                "peak_context_tokens": 92336,
                "steps": 41,
                "requests": 42
              },
              "patch": {
                "bytes": 33220,
                "sha256": "4e0099d8be5cc3ca857516e9adfc9690499770cdb4d6db68e2330dce7ca6c05f"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "kombu-single-active-consumer-priority",
          "official_prompt": {
            "title": "Add single-active-consumer priority and cancel tracking to virtual transports",
            "description": "Add single-active-consumer semantics, consumer priority selection, cancel notifications, and lifecycle tracking to virtual transports.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/kombu-single-active-consumer-priority",
            "instruction_sha256": "c40c52ac1ec444168f42dbf58babb9d92360d1cff4b5d71cb3abd388832b2cf2",
            "task_toml_sha256": "9291e2b4b0a6dbbc83a35658943e2c9037bce4170014481b798cbd80b2e2ebff",
            "language": "python",
            "category": "feature_request"
          },
          "repository": "https://github.com/celery/kombu",
          "base_revision": "3c5c1bd86376ee73d52a4cc770bdaeab15bbc2f3",
          "runs": [
            {
              "run_id": "tonight-20261027/kombu-single-active-consumer-pri__zihagvt",
              "task_id": "kombu-single-active-consumer-priority",
              "campaign": "tonight-20261027",
              "outcome": "pass",
              "score": {
                "feature": {
                  "passed": 85,
                  "total": 85
                },
                "preservation": {
                  "passed": 1421,
                  "total": 1421
                },
                "reward": 1
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "context": 196608,
                "engine": "Strata 0.1.48.1 HC_Q8",
                "output_policy": "native remaining context",
                "agent_seconds_limit": 2700
              },
              "timing_seconds": {
                "environment_setup": 27.193401,
                "agent_setup": 0.297871,
                "agent_execution": 646.223202,
                "verifier": 49.377818
              },
              "usage": {
                "input_tokens": 2472754,
                "cached_tokens": 2439192,
                "output_tokens": 30632,
                "peak_context_tokens": 69569,
                "steps": 50,
                "requests": 51
              },
              "patch": {
                "bytes": 26255,
                "sha256": "8ca66817961cc56c63ab318c1118b24e4e615373fd0bb8dee7bcd7b30111a755"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "tengo-callable-instance-isolation",
          "official_prompt": {
            "title": "Fix isolated Go-side calls for Tengo callables and closures",
            "description": "Enable Go-side invocation of exported Tengo functions and closures while preserving runtime context and isolating compiled-instance state.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/tengo-callable-instance-isolation",
            "instruction_sha256": "ea2d12e6c683a868e29d90973d13da585a15f69c633a56570bf4492fb5091ec5",
            "task_toml_sha256": "5d9da4f19acad54ac5262742ecc6856ccc763e6902717a858e8379bfc5a0ed90",
            "language": "go",
            "category": "bugfix"
          },
          "repository": "https://github.com/d5/tengo",
          "base_revision": "3cad0da7a51b1206c6f01e3f4fbb44b976d5275c",
          "runs": [
            {
              "run_id": "tonight-20261027/tengo-callable-instance-isolatio__DcuM4At",
              "task_id": "tengo-callable-instance-isolation",
              "campaign": "tonight-20261027",
              "outcome": "pass",
              "score": {
                "feature": {
                  "passed": 23,
                  "total": 23
                },
                "preservation": {
                  "passed": 122,
                  "total": 122
                },
                "reward": 1
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "context": 196608,
                "engine": "Strata 0.1.48.1 HC_Q8",
                "output_policy": "native remaining context",
                "agent_seconds_limit": 2700
              },
              "timing_seconds": {
                "environment_setup": 25.31983,
                "agent_setup": 0.295153,
                "agent_execution": 1800.899705,
                "verifier": 26.517798
              },
              "usage": {
                "input_tokens": 7930020,
                "cached_tokens": 7739465,
                "output_tokens": 93136,
                "peak_context_tokens": 156223,
                "steps": 81,
                "requests": 82
              },
              "patch": {
                "bytes": 37177,
                "sha256": "0bb19a8150734cfab4e7affd71cef961be4c69df25202424bc572fcff10b0f44"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "optique-conditional-option-dependencies",
          "official_prompt": {
            "title": "Add conditional option dependencies to Optique",
            "description": "Add conditional dependencies so options can require or hide based on other option presence or values.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/optique-conditional-option-dependencies",
            "instruction_sha256": "bfe35f35e82efdeba62d75a21df8687029ce5c0e50bddba5327e4285cc5cc41b",
            "task_toml_sha256": "21fdf725891f2c73a44723242e397fd0333800c2209f76f83eb8ed2d357ec273",
            "language": "typescript",
            "category": "feature_request"
          },
          "repository": "https://github.com/dahlia/optique",
          "base_revision": "14bbe4efc7ded67932771b9ca18d9d637bb4cf27",
          "runs": [
            {
              "run_id": "tonight-20261027/optique-conditional-option-depen__pguqpwb",
              "task_id": "optique-conditional-option-dependencies",
              "campaign": "tonight-20261027",
              "outcome": "graded-model-failure",
              "score": {
                "feature": {
                  "passed": 35,
                  "total": 36
                },
                "preservation": {
                  "passed": 2034,
                  "total": 2034
                },
                "reward": 0
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "context": 196608,
                "engine": "Strata 0.1.48.1 HC_Q8",
                "output_policy": "native remaining context",
                "agent_seconds_limit": 2700
              },
              "timing_seconds": {
                "environment_setup": 39.897904,
                "agent_setup": 0.295574,
                "agent_execution": 1698.608468,
                "verifier": 23.544547
              },
              "usage": {
                "input_tokens": 8191828,
                "cached_tokens": 7972739,
                "output_tokens": 84908,
                "peak_context_tokens": 154601,
                "steps": 113,
                "requests": 114
              },
              "patch": {
                "bytes": 53002,
                "sha256": "b8dd68214e5a378ea74e5e2377ea6f8ff71c7c7832f6a822ffe6e74424d265e2"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              }
            }
          ],
          "frontier": [],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "kysely-window-grouping-helpers",
          "official_prompt": {
            "title": "Add grouping-set and window-frame SQL helpers",
            "description": "Add grouped aggregation clauses, window frame builders, null-handling modifiers, and related SQL function helpers.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/kysely-window-grouping-helpers",
            "instruction_sha256": "7b56ec113b59dc9386b1ea5c7c6d252c6dbcef333b7740744cfbd454eacfa19d",
            "task_toml_sha256": "965f2e0b07ea9ff508c6cac0712c853c6b1fbffba31955c8bf2e628fca013ce5",
            "language": "typescript",
            "category": "feature_request"
          },
          "repository": "https://github.com/kysely-org/kysely",
          "base_revision": "91cf3733b2a419f5b17dff118cedb7052ab5300d",
          "runs": [
            {
              "run_id": "tonight-20261027/kysely-window-grouping-helpers__b4Z4seR",
              "task_id": "kysely-window-grouping-helpers",
              "campaign": "tonight-20261027",
              "outcome": "pass",
              "score": {
                "feature": {
                  "passed": 254,
                  "total": 254
                },
                "preservation": {
                  "passed": 22,
                  "total": 22
                },
                "reward": 1
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "context": 196608,
                "engine": "Strata 0.1.48.1 HC_Q8",
                "output_policy": "native remaining context",
                "agent_seconds_limit": 2700,
                "compaction": {
                  "auto": true,
                  "prune": false,
                  "reserved": 40000
                },
                "mtp": 4
              },
              "timing_seconds": {
                "environment_setup": 34.564512,
                "agent_setup": 0.286212,
                "agent_execution": 1304.008379,
                "verifier": 40.151811
              },
              "usage": {
                "input_tokens": 9343696,
                "cached_tokens": 9287643,
                "output_tokens": 67179,
                "peak_context_tokens": 122864,
                "steps": 118,
                "requests": 119
              },
              "patch": {
                "bytes": 80851,
                "sha256": "02879ceb31407b5fba9bbe5150ed98c9c0e26da3dc15cae7f2b5e1aa28ab978e"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              },
              "started_at": "2026-10-28T06:11:00.409016Z",
              "finished_at": "2026-10-28T06:34:11.985205Z"
            }
          ],
          "frontier": [],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "opa-rego-rule-profiling",
          "official_prompt": {
            "title": "Add rule evaluation profiling to Rego",
            "description": "Add opt-in per-rule evaluation profiling to Rego results with profiling stats and diff helpers.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/opa-rego-rule-profiling",
            "instruction_sha256": "7892902b74c649f6cb76caf6d311b880ecff88179e6b5afaceedaa143160ae87",
            "task_toml_sha256": "1584034120b0c598b342c56f1a88c95ca2d7d16be64e2e790ce229a6055a2c28",
            "language": "go",
            "category": "feature_request"
          },
          "repository": "https://github.com/open-policy-agent/opa",
          "base_revision": "1ac64ef1a57a531c2723c59848890b88e816d777",
          "runs": [
            {
              "run_id": "tonight-20261027/opa-rego-rule-profiling__8jeoKdt",
              "task_id": "opa-rego-rule-profiling",
              "campaign": "tonight-20261027",
              "outcome": "pass",
              "score": {
                "feature": {
                  "passed": 25,
                  "total": 25
                },
                "preservation": {
                  "passed": 6,
                  "total": 6
                },
                "reward": 1
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "context": 196608,
                "engine": "Strata 0.1.48.1 HC_Q8",
                "output_policy": "native remaining context",
                "agent_seconds_limit": 2700,
                "compaction": {
                  "auto": true,
                  "prune": false,
                  "reserved": 40000
                },
                "mtp": 4
              },
              "timing_seconds": {
                "environment_setup": 41.689212,
                "agent_setup": 0.289415,
                "agent_execution": 1873.441408,
                "verifier": 47.693179
              },
              "usage": {
                "input_tokens": 12013561,
                "cached_tokens": 11809543,
                "output_tokens": 88358,
                "peak_context_tokens": 156540,
                "steps": 125,
                "requests": 126
              },
              "patch": {
                "bytes": 46578,
                "sha256": "3f96a52a9f866e4b8c4850ab6b66c1a990b1d7e9aafd837bc27945c2ab6643ed"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              },
              "started_at": "2026-10-28T06:34:22.084367Z",
              "finished_at": "2026-10-28T07:07:18.041269Z"
            }
          ],
          "frontier": [],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "koota-composite-trait-aspects",
          "official_prompt": {
            "title": "Add composite trait aspects to Koota",
            "description": "Add createAspect and aspect-aware query operations that merge constituent traits and propagate writes and lifecycle events.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/koota-composite-trait-aspects",
            "instruction_sha256": "300c70ea3b277bf799c272cb7415368bb631856c5a89bee1e7cac474f5810c25",
            "task_toml_sha256": "fa101fd31b866c79a5d14188e5d1a8e84d9a770703f4c5ac5c983eff4a599ee7",
            "language": "typescript",
            "category": "feature_request"
          },
          "repository": "https://github.com/pmndrs/koota",
          "base_revision": "9c434858b2b522002f8c5eb4a554fa8836a7cf3c",
          "runs": [
            {
              "run_id": "tonight-20261027/koota-composite-trait-aspects__3dm73dq",
              "task_id": "koota-composite-trait-aspects",
              "campaign": "tonight-20261027",
              "outcome": "pass",
              "score": {
                "feature": {
                  "passed": 51,
                  "total": 51
                },
                "preservation": {
                  "passed": 172,
                  "total": 172
                },
                "reward": 1
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "context": 196608,
                "engine": "Strata 0.1.48.1 HC_Q8",
                "output_policy": "native remaining context",
                "agent_seconds_limit": 2700,
                "compaction": {
                  "auto": true,
                  "prune": false,
                  "reserved": 40000
                },
                "mtp": 4
              },
              "timing_seconds": {
                "environment_setup": 30.899988,
                "agent_setup": 0.306779,
                "agent_execution": 2101.744903,
                "verifier": 23.21005
              },
              "usage": {
                "input_tokens": 12328144,
                "cached_tokens": 12130356,
                "output_tokens": 109946,
                "peak_context_tokens": 156886,
                "steps": 145,
                "requests": 146
              },
              "patch": {
                "bytes": 123316,
                "sha256": "99c54220683c940ee0b0599b940259333d8cd78e13ab2875455ce584388b21da"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              },
              "started_at": "2026-10-28T08:17:10.795847Z",
              "finished_at": "2026-10-28T08:53:19.519623Z"
            }
          ],
          "frontier": [],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "obsidian-linter-link-format-conversion",
          "official_prompt": {
            "title": "Add link format conversion between wiki and markdown syntax",
            "description": "Add a Link Style rule that converts between Obsidian wiki links or embeds and markdown links or images while preserving ignored regions and edge cases.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/obsidian-linter-link-format-conversion",
            "instruction_sha256": "c54e9a11b92254161afeb32cbce31ada65524854df372f889c00f5d9ee8962a9",
            "task_toml_sha256": "52a326cb1d708909e778398b71920e0c43f86ef533f14f01a8f5d117e2dc7951",
            "language": "typescript",
            "category": "feature_request"
          },
          "repository": "https://github.com/platers/obsidian-linter",
          "base_revision": "6393b3ab32a2ace1fc24d4b0f5e0f13a179c874f",
          "runs": [
            {
              "run_id": "tonight-20261027/obsidian-linter-link-format-conv__g7LnXvY",
              "task_id": "obsidian-linter-link-format-conversion",
              "campaign": "tonight-20261027",
              "outcome": "graded-model-failure",
              "score": {
                "feature": {
                  "passed": 0,
                  "total": 60
                },
                "preservation": {
                  "passed": 1131,
                  "total": 1131
                },
                "reward": 0
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "context": 196608,
                "engine": "Strata 0.1.48.1 HC_Q8",
                "output_policy": "native remaining context",
                "agent_seconds_limit": 2700,
                "compaction": {
                  "auto": true,
                  "prune": false,
                  "reserved": 40000
                },
                "mtp": 4
              },
              "timing_seconds": {
                "environment_setup": 28.134449,
                "agent_setup": 0.293336,
                "agent_execution": 893.684247,
                "verifier": 32.508323
              },
              "usage": {
                "input_tokens": 3396644,
                "cached_tokens": 3356523,
                "output_tokens": 45083,
                "peak_context_tokens": 82545,
                "steps": 70,
                "requests": 71
              },
              "patch": {
                "bytes": 29933,
                "sha256": "c600fabf60c269a3707b6eac19da9f69c1610061c66ee41049bdb654ccc40418"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              },
              "started_at": "2026-10-28T08:00:49.743737Z",
              "finished_at": "2026-10-28T08:16:57.121149Z"
            }
          ],
          "frontier": [],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "prometheus-typed-label-sorting",
          "official_prompt": {
            "title": "Fix PromQL label sorting across typed and untyped values",
            "description": "PromQL label sorting must order mixed typed and untyped label values with stable typed comparison rules.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/prometheus-typed-label-sorting",
            "instruction_sha256": "02ffc5959762f9358d2b6a046c808e1468cc5afdf52bdcd7c36c6020b8b748df",
            "task_toml_sha256": "ee7b6fe19ee8477f8a2eda14bf36da1eba3ab41a52e475a9f7f5118e23e4b1f1",
            "language": "go",
            "category": "bugfix"
          },
          "repository": "https://github.com/prometheus/prometheus",
          "base_revision": "8b25b26a7653d9c7444f217a7f2ae9b327bda921",
          "runs": [
            {
              "run_id": "tonight-20261027/prometheus-typed-label-sorting__cu68Mhy",
              "task_id": "prometheus-typed-label-sorting",
              "campaign": "tonight-20261027",
              "outcome": "graded-model-failure",
              "score": {
                "feature": {
                  "passed": 0,
                  "total": 17
                },
                "preservation": {
                  "passed": 28,
                  "total": 28
                },
                "reward": 0
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "context": 196608,
                "engine": "Strata 0.1.48.1 HC_Q8",
                "output_policy": "native remaining context",
                "agent_seconds_limit": 2700,
                "compaction": {
                  "auto": true,
                  "prune": false,
                  "reserved": 40000
                },
                "mtp": 4
              },
              "timing_seconds": {
                "environment_setup": 47.702286,
                "agent_setup": 0.296196,
                "agent_execution": 735.087552,
                "verifier": 59.253862
              },
              "usage": {
                "input_tokens": 940567,
                "cached_tokens": 928854,
                "output_tokens": 38803,
                "peak_context_tokens": 50269,
                "steps": 31,
                "requests": 32
              },
              "patch": {
                "bytes": 25035,
                "sha256": "354d77b203507da14e759f72c89f7dc8d8f8b31a5a6ad60ba2dff198a276c976"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              },
              "started_at": "2026-10-28T08:53:33.211974Z",
              "finished_at": "2026-10-28T09:07:48.622702Z"
            }
          ],
          "frontier": [],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "psd-tools-blend-range-api",
          "official_prompt": {
            "title": "Add typed blend range access and blend-if compositing",
            "description": "Add typed blend range objects, persist them on layers, and apply blend-if during compositing.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/psd-tools-blend-range-api",
            "instruction_sha256": "be619e943671da3bb1e9855c0ce939d54491c5cce4a45553403cf026d33255ce",
            "task_toml_sha256": "e929855b9ee6d1874df3f9a628a8ba60243c06bb7c97d40a2f60286ef5c289e9",
            "language": "python",
            "category": "feature_request"
          },
          "repository": "https://github.com/psd-tools/psd-tools",
          "base_revision": "c5e03189188daa3c5589326a9d74506d7dc48bc9",
          "runs": [
            {
              "run_id": "tonight-20261027/psd-tools-blend-range-api__CPJGVpD",
              "task_id": "psd-tools-blend-range-api",
              "campaign": "tonight-20261027",
              "outcome": "pass",
              "score": {
                "feature": {
                  "passed": 45,
                  "total": 45
                },
                "preservation": {
                  "passed": 979,
                  "total": 979
                },
                "reward": 1
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "context": 196608,
                "engine": "Strata 0.1.48.1 HC_Q8",
                "output_policy": "native remaining context",
                "agent_seconds_limit": 2700,
                "compaction": {
                  "auto": true,
                  "prune": false,
                  "reserved": 40000
                },
                "mtp": 4
              },
              "timing_seconds": {
                "environment_setup": 27.097086,
                "agent_setup": 0.300097,
                "agent_execution": 966.818405,
                "verifier": 29.27253
              },
              "usage": {
                "input_tokens": 4586892,
                "cached_tokens": 4540213,
                "output_tokens": 47422,
                "peak_context_tokens": 91362,
                "steps": 78,
                "requests": 79
              },
              "patch": {
                "bytes": 35871,
                "sha256": "4e7b21abaadbcd10392ab62dbe336b25655d9fb773c75a59445d009d9bf4670d"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              },
              "started_at": "2026-10-28T09:07:54.436296Z",
              "finished_at": "2026-10-28T09:25:10.469905Z"
            }
          ],
          "frontier": [],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        },
        {
          "task_id": "quill-shared-toolbar-focus",
          "official_prompt": {
            "title": "Reuse one toolbar across multiple Quill editors",
            "description": "Allow multiple Quill editors to share one toolbar container while routing toolbar actions and UI state to the currently active editor.",
            "description_source": "official task prompt",
            "url": "https://deepswe.datacurve.ai/data/v1.1/tasks/quill-shared-toolbar-focus",
            "instruction_sha256": "a983670bc6aaee7de85eb8aeef1a4f1db44788ecc36905a0bccbec469e0c73cd",
            "task_toml_sha256": "01df5b9915d64e073900c2a00da5a71639614b3baa24f9d5f83bd401740073ac",
            "language": "typescript",
            "category": "feature_request"
          },
          "repository": "https://github.com/slab/quill",
          "base_revision": "539cbffd0a13b18e9c65eb84dd35e6596e403158",
          "runs": [
            {
              "run_id": "tonight-20261027/quill-shared-toolbar-focus__bCXFuyX",
              "task_id": "quill-shared-toolbar-focus",
              "campaign": "tonight-20261027",
              "outcome": "pass",
              "score": {
                "feature": {
                  "passed": 13,
                  "total": 13
                },
                "preservation": {
                  "passed": 22,
                  "total": 22
                },
                "reward": 1
              },
              "configuration": {
                "model": "qwen3.8-flash-next-iq3_s",
                "reasoning_effort": "medium",
                "max_tokens": 196608,
                "temperature": 1.0,
                "top_p": 0.95,
                "top_k": 20,
                "min_p": 0.0,
                "presence_penalty": 0.0,
                "repetition_penalty": 1.0,
                "context": 196608,
                "engine": "Strata 0.1.48.1 HC_Q8",
                "output_policy": "native remaining context",
                "agent_seconds_limit": 2700,
                "compaction": {
                  "auto": true,
                  "prune": false,
                  "reserved": 40000
                },
                "mtp": 4
              },
              "timing_seconds": {
                "environment_setup": 38.971619,
                "agent_setup": 0.293128,
                "agent_execution": 1557.593723,
                "verifier": 21.603668
              },
              "usage": {
                "input_tokens": 6712673,
                "cached_tokens": 6642944,
                "output_tokens": 80159,
                "peak_context_tokens": 148191,
                "steps": 69,
                "requests": 70
              },
              "patch": {
                "bytes": 36677,
                "sha256": "55709a5fc76c0949f9fa9aa3db2de2161a7c60ae2e1f4baaffe881ddeac8aba2"
              },
              "verification": {
                "status": "official-scored",
                "verifier": "DeepSWE v1.1 official verifier",
                "evidence": "verified campaign state and official verifier result"
              },
              "started_at": "2026-10-28T09:25:15.776701Z",
              "finished_at": "2026-10-28T09:52:26.842044Z"
            }
          ],
          "frontier": [],
          "frontier_note": "Published per-task evidence from a different harness; not this run."
        }
      ]
    }
  ]
}