{
  "schema_version": "0.7",
  "generation_sampled": true,
  "skill": {
    "name": "documentation-and-adrs",
    "version": "0.0.0",
    "content_hash": "b67a9f07ed1057963d3d8ab04bb070ee74d93174eee48b563aa235057005c158",
    "tokens": 2454
  },
  "suite": {
    "format": "agentskills.io/evals",
    "suite_hash": "6cce54e7d9d0fc923eda5499c75fcc68974b3d8714b2a726e2ad1d77e4b5acc0",
    "case_count": 1,
    "canary": "86e8fbf6-9c40-ec04-dacc-03e730017455"
  },
  "run": {
    "model_id": "claude-opus-5-5",
    "model_release_date": null,
    "provider": "anthropic",
    "surface": "claude-cli",
    "runner_version": "0.11.2",
    "date_utc": "2026-09-24T07:15:14.215Z",
    "registry": "registered",
    "transcripts": "hashes-only",
    "judge": {
      "samples": 3,
      "temperature": null,
      "sampling": "surface-controlled",
      "surface": "claude-cli",
      "model_id": "claude-opus-5",
      "prompt_template_hash": "82586d1e44f84df2b8ce5f18ac777866465cb5ad9686503c93d9776c93c344ed"
    },
    "answered_by": {
      "kind": "model",
      "attested": true,
      "reported_model": "claude-opus-5-5",
      "reported_models": [
        "claude-opus-5",
        "claude-opus-5-5"
      ],
      "isolation": "same-user"
    },
    "pricing_snapshot": {
      "frozen_at": "2026-09-24T07:15:14.215Z",
      "source": "config/models.json",
      "currency": "USD",
      "models": {
        "claude-opus-5-5": {
          "input_per_mtok": 4,
          "output_per_mtok": 20,
          "registered": true
        },
        "claude-opus-5": {
          "input_per_mtok": 5,
          "output_per_mtok": 25,
          "registered": true
        }
      },
      "note": "Prices frozen at run time. Every derived cost in this receipt is computed from THIS snapshot, never from the live registry, so the receipt keeps its meaning when registry prices later change."
    }
  },
  "results": {
    "cases": [
      {
        "id": "comment-intent-not-implementation",
        "mode": "with_skill",
        "outcome": "borderline",
        "score": 0.69,
        "mean": 0.69,
        "stddev": 0.14709,
        "samples": [
          0.55,
          0.55,
          0.6
        ],
        "generation_hash": "f4ed0167d112d4c095315111c9764efddc5e05344aa2e86d52dd4b8cc1f723e2",
        "judge_sample_hashes": [
          "b1c2401229e0a20702ba6e6679f9762b2f88b108db4a61cf2b8bd338200a8f10",
          "89d0c55598259608939afcdc984a4dbc17f937dc418b6f4b3bcc3b80d7e09a57",
          "209c7070f5c19de213c6823d42fe8ad3d1cc9db097ab8781370705d6e1009fcc"
        ],
        "threshold": 0.7,
        "reason": "Meets (a) and (b) fully and (d) largely via the side-effect doc comment, but leaves the TODO unimplemented with an admittedly placeholder reason, and '10% gold-tier discount' restates the what.",
        "judge": {
          "model_id": "claude-opus-5",
          "rubric_hash": "b38f01e6671fc0b28a94de550fe70f879134a3168b35e5b1fb53a9a6773478b0"
        },
        "judge_usage": {
          "input_tokens": 18918,
          "output_tokens": 2390,
          "cached_tokens": 15864,
          "wall_ms": 37565
        },
        "generation": {
          "n_planned": 3,
          "n_drawn": 6,
          "n_measured": 6,
          "n_unmeasured": 0,
          "stopping_reason": "stabilised",
          "mean": 0.69,
          "sd": 0.14709,
          "judge_sd_mean": 0.038112,
          "variance_ratio": 3.859414,
          "variance_ratio_unavailable": null,
          "n_truncated": 0,
          "draws": [
            {
              "draw_index": 0,
              "generation_hash": "c7276cb74f981e1cc536279f25197b31c4aefb2fdc0dd917e187f8ebff785457",
              "status": "measured",
              "samples": [
                0.6,
                0.5,
                0.45
              ],
              "judge_sample_hashes": [
                "169645e23fd76454a833ad8dd0706823902a1939410cdfdf44b3fcdab3696091",
                "28f0379cdbda0726b7d98d58f707394e19c1553231f7838f75eeb7fa3daf249d",
                "e2ad68fe01ae225455e6a118893a308fb03793ae1a7fd180540d50211c404fa9"
              ],
              "mean": 0.516667,
              "stddev": 0.076376,
              "stop_reason": "end_turn",
              "truncated": false,
              "reported_model": "claude-opus-5-5",
              "usage": {
                "input_tokens": 7187,
                "output_tokens": 840,
                "cached_tokens": 540,
                "wall_ms": 10773
              },
              "judge_usage": {
                "input_tokens": 18816,
                "output_tokens": 2575,
                "cached_tokens": 15796,
                "wall_ms": 41544
              }
            },
            {
              "draw_index": 1,
              "generation_hash": "283e07542179f8f516e1486ff00010a649c9a06f8024bd1c26b819605dab2901",
              "status": "measured",
              "samples": [
                0.83,
                0.82,
                0.85
              ],
              "judge_sample_hashes": [
                "b8611b7d539ba18475eb23a0b94d3b8a5c8f1a5b50666262e36f950b2875aa06",
                "9a5203f5a4162aaaf0b67b22a6a6bdfac8c549382b92f317330fc5bde6af2f16",
                "5fad7cbe40938790c714837bd5ef32f29c23fc07de35d495a4855a5c3030d323"
              ],
              "mean": 0.833333,
              "stddev": 0.015275,
              "stop_reason": "end_turn",
              "truncated": false,
              "reported_model": "claude-opus-5-5",
              "usage": {
                "input_tokens": 7187,
                "output_tokens": 808,
                "cached_tokens": 7185,
                "wall_ms": 10607
              },
              "judge_usage": {
                "input_tokens": 18786,
                "output_tokens": 1953,
                "cached_tokens": 15776,
                "wall_ms": 29738
              }
            },
            {
              "draw_index": 2,
              "generation_hash": "b2d75929af9576560652cf901b7a7234f8896dca79bd19d041fda0dbcada8bb4",
              "status": "measured",
              "samples": [
                0.6,
                0.6,
                0.6
              ],
              "judge_sample_hashes": [
                "9efb292528c65030737cf6192ac50219459a4ff748dcdffed0a6224e35669fdd",
                "eb6c28f4e157c0ba3ddd268437d50eefde2215bcc11c3526df47ee3bd94ae9ac",
                "c947922d78e0710e9ecf86fecca5e01580c3d3107b354abc6ef4dbe1293aa638"
              ],
              "mean": 0.6,
              "stddev": 0,
              "stop_reason": "end_turn",
              "truncated": false,
              "reported_model": "claude-opus-5-5",
              "usage": {
                "input_tokens": 7187,
                "output_tokens": 829,
                "cached_tokens": 7185,
                "wall_ms": 10878
              },
              "judge_usage": {
                "input_tokens": 18570,
                "output_tokens": 1690,
                "cached_tokens": 15632,
                "wall_ms": 26586
              }
            },
            {
              "draw_index": 3,
              "generation_hash": "7a9065550fdd632dbd133d985a10df4888dc7736b1a8b01783853d4ddf87b9b1",
              "status": "measured",
              "samples": [
                0.86,
                0.87,
                0.85
              ],
              "judge_sample_hashes": [
                "4f2fa9d7901cac457e819f02610dde1799f3756f16d0e8aa7b5751fd4e28f10b",
                "1fbd9f62b9a5cf3a5f8e1c4b8d07ebaa4a1d8e3053c9cca7582c0e35c3e76049",
                "2d04870ff571357f2d21fd1c72fdd96064841da5c03cff1cb7fc4571cfb8c4b4"
              ],
              "mean": 0.86,
              "stddev": 0.01,
              "stop_reason": "end_turn",
              "truncated": false,
              "reported_model": "claude-opus-5-5",
              "usage": {
                "input_tokens": 7187,
                "output_tokens": 936,
                "cached_tokens": 7185,
                "wall_ms": 12048
              },
              "judge_usage": {
                "input_tokens": 19026,
                "output_tokens": 1337,
                "cached_tokens": 15936,
                "wall_ms": 23129
              }
            },
            {
              "draw_index": 4,
              "generation_hash": "4b5b0738272a3852070841c4da086f8f3e71a08cf8daf7629aa8a2e45101bad6",
              "status": "measured",
              "samples": [
                0.82,
                0.82,
                0.65
              ],
              "judge_sample_hashes": [
                "6d21dfb33ea3d39d18f9dc49a75522dadff0b0145d1dbbc914154ecaa4a68a42",
                "19eaef0eef751ba2f4f2b5ecef47dd76765a5d9de17a52e39be8841fdf79c017",
                "ec8eef520535cc5ac45f7689c49dbbbd174a3726f25a3a809b8c05427f2d73ef"
              ],
              "mean": 0.763333,
              "stddev": 0.09815,
              "stop_reason": "end_turn",
              "truncated": false,
              "reported_model": "claude-opus-5-5",
              "usage": {
                "input_tokens": 7187,
                "output_tokens": 889,
                "cached_tokens": 7185,
                "wall_ms": 10835
              },
              "judge_usage": {
                "input_tokens": 19065,
                "output_tokens": 2621,
                "cached_tokens": 15962,
                "wall_ms": 44039
              }
            },
            {
              "draw_index": 5,
              "generation_hash": "f4ed0167d112d4c095315111c9764efddc5e05344aa2e86d52dd4b8cc1f723e2",
              "status": "measured",
              "samples": [
                0.55,
                0.55,
                0.6
              ],
              "judge_sample_hashes": [
                "b1c2401229e0a20702ba6e6679f9762b2f88b108db4a61cf2b8bd338200a8f10",
                "89d0c55598259608939afcdc984a4dbc17f937dc418b6f4b3bcc3b80d7e09a57",
                "209c7070f5c19de213c6823d42fe8ad3d1cc9db097ab8781370705d6e1009fcc"
              ],
              "mean": 0.566667,
              "stddev": 0.028868,
              "stop_reason": "end_turn",
              "truncated": false,
              "reported_model": "claude-opus-5-5",
              "usage": {
                "input_tokens": 7187,
                "output_tokens": 841,
                "cached_tokens": 7185,
                "wall_ms": 10882
              },
              "judge_usage": {
                "input_tokens": 18918,
                "output_tokens": 2390,
                "cached_tokens": 15864,
                "wall_ms": 37565
              }
            }
          ]
        }
      },
      {
        "id": "comment-intent-not-implementation",
        "mode": "baseline",
        "outcome": "fail",
        "score": 0.573333,
        "mean": 0.573333,
        "stddev": 0.017638,
        "samples": [
          0.6,
          0.55,
          0.55
        ],
        "generation_hash": "294a63179dd624e32b3121cb9b74d9eb7d738ee60ddbccc783713c2cb8dad2db",
        "judge_sample_hashes": [
          "f69c697d4aa938fca794714ffb00701d4f3fec97654557ff6b1e2b6e29aef100",
          "d0982493ce601c17eb56f24c6014e4cc4d9bf7b6d316a6862ad864221eb75688",
          "c883b332cadd267898b76be9d7600064e2e39d9ea2bccbbb0b582377706b5716"
        ],
        "threshold": 0.7,
        "reason": "Meets (a), (b), and largely (d) with genuine intent comments, but leaves an unimplemented TODO rather than handling or removing it; gold-tier comment partly restates the what.",
        "judge": {
          "model_id": "claude-opus-5",
          "rubric_hash": "b38f01e6671fc0b28a94de550fe70f879134a3168b35e5b1fb53a9a6773478b0"
        },
        "judge_usage": {
          "input_tokens": 18366,
          "output_tokens": 2332,
          "cached_tokens": 15496,
          "wall_ms": 33530
        },
        "generation": {
          "n_planned": 3,
          "n_drawn": 3,
          "n_measured": 3,
          "n_unmeasured": 0,
          "stopping_reason": "min_reached",
          "mean": 0.573333,
          "sd": 0.017638,
          "judge_sd_mean": 0.019245,
          "variance_ratio": 0.916498,
          "variance_ratio_unavailable": null,
          "n_truncated": 0,
          "draws": [
            {
              "draw_index": 0,
              "generation_hash": "c4fa9e2bf62d5df93b2ba7f86688d91002704b7d97726dfb3da799c082e36046",
              "status": "measured",
              "samples": [
                0.55,
                0.58,
                0.55
              ],
              "judge_sample_hashes": [
                "7255266d0599f96746e614bd57abeab98f8b331ba2e9a06e71d16b3890d59ab5",
                "f57e3132add4fae844ca2edde86c2ef20dc6a332fc09542fbdec14730c85c4b8",
                "6ac06cae91ca7a06d1afdeac5fa888645d30c15de0fe71ce47dfeb80da77c051"
              ],
              "mean": 0.56,
              "stddev": 0.017321,
              "stop_reason": "end_turn",
              "truncated": false,
              "reported_model": "claude-opus-5-5",
              "usage": {
                "input_tokens": 3724,
                "output_tokens": 528,
                "cached_tokens": 2145,
                "wall_ms": 7973
              },
              "judge_usage": {
                "input_tokens": 18267,
                "output_tokens": 1847,
                "cached_tokens": 15430,
                "wall_ms": 29024
              }
            },
            {
              "draw_index": 1,
              "generation_hash": "c2248b92091f2d510922c39791ea89516068f04782b71c69531a54ddfda88f90",
              "status": "measured",
              "samples": [
                0.58,
                0.6,
                0.6
              ],
              "judge_sample_hashes": [
                "39f73fc908f3d1a3ff82f1b85e1e27c3361e821496a5c238ca063445c39944f3",
                "c27772cbe700e8868344461447b32bc03cdfbd23b52d7c20360a97964eeca4eb",
                "04d1e585af801e2908b3706d2386f77ebf903f5c2f1c1a83e2928ce3389fc4e1"
              ],
              "mean": 0.593333,
              "stddev": 0.011547,
              "stop_reason": "end_turn",
              "truncated": false,
              "reported_model": "claude-opus-5-5",
              "usage": {
                "input_tokens": 3724,
                "output_tokens": 721,
                "cached_tokens": 3722,
                "wall_ms": 9540
              },
              "judge_usage": {
                "input_tokens": 18534,
                "output_tokens": 1753,
                "cached_tokens": 15608,
                "wall_ms": 29571
              }
            },
            {
              "draw_index": 2,
              "generation_hash": "294a63179dd624e32b3121cb9b74d9eb7d738ee60ddbccc783713c2cb8dad2db",
              "status": "measured",
              "samples": [
                0.6,
                0.55,
                0.55
              ],
              "judge_sample_hashes": [
                "f69c697d4aa938fca794714ffb00701d4f3fec97654557ff6b1e2b6e29aef100",
                "d0982493ce601c17eb56f24c6014e4cc4d9bf7b6d316a6862ad864221eb75688",
                "c883b332cadd267898b76be9d7600064e2e39d9ea2bccbbb0b582377706b5716"
              ],
              "mean": 0.566667,
              "stddev": 0.028868,
              "stop_reason": "end_turn",
              "truncated": false,
              "reported_model": "claude-opus-5-5",
              "usage": {
                "input_tokens": 3724,
                "output_tokens": 587,
                "cached_tokens": 3722,
                "wall_ms": 8364
              },
              "judge_usage": {
                "input_tokens": 18366,
                "output_tokens": 2332,
                "cached_tokens": 15496,
                "wall_ms": 33530
              }
            }
          ]
        }
      }
    ],
    "aggregates": {
      "with_skill": {
        "case_count": 1,
        "pass_count": 0,
        "borderline_count": 1,
        "mean_score": 0.69,
        "stddev": null
      },
      "baseline": {
        "case_count": 1,
        "pass_count": 0,
        "borderline_count": 0,
        "mean_score": 0.573333,
        "stddev": null
      },
      "band_rule": "per arm: the sample standard deviation (n-1) of the per-case means across the included cases; the comparison band is the quadrature sum of the two arms; null where an arm has fewer than two included cases"
    }
  },
  "comparison": {
    "with_skill_score": 0.69,
    "baseline_score": 0.573333,
    "delta": 0.116667,
    "delta_uncertainty": null,
    "delta_uncertainty_unavailable": "single_case"
  },
  "verification_level": "TESTED",
  "receipt_hash": "7bb16fde55abfd55c4bc8a923ad6390a712318117a711960c0b5c68e81741d13",
  "economics": {
    "basis": "metered-equivalent",
    "surface": "claude-cli",
    "with_skill": {
      "call_count": 6,
      "mean_input_tokens": 7187,
      "mean_output_tokens": 857.17,
      "mean_cost_usd_per_call": 0.045891,
      "median_wall_ms": 10856.5,
      "wall_ms_p25": 10773,
      "wall_ms_p75": 10882,
      "wall_ms_iqr": 109
    },
    "baseline": {
      "call_count": 3,
      "mean_input_tokens": 3724,
      "mean_output_tokens": 612,
      "mean_cost_usd_per_call": 0.027136,
      "median_wall_ms": 8364,
      "wall_ms_p25": null,
      "wall_ms_p75": null,
      "wall_ms_iqr": null
    },
    "skill_incremental_cost_usd_per_call": 0.018755,
    "skill_incremental_cost_usd_per_1k_calls": 18.755,
    "output_tokens_delta": 245.17,
    "median_wall_ms_delta": 2492.5,
    "judge_excluded": true,
    "judge_overhead": {
      "note": "Measurement overhead imposed by Driftproof, NOT a cost of running the skill. Excluded from every field above.",
      "total_cost_usd": 0.30447,
      "case_rows_measured": 2
    },
    "notes": {
      "absolute_cost": "Absolute per-call cost on a CLI surface includes a fixed harness preamble (observed ~25k input tokens on claude-cli, ~11k on codex) that we do not control. It is identical in the with-skill and baseline arms, so it cancels in the Δ figures — read the incremental columns, not the absolute ones.",
      "cache_pricing": "Cached input tokens are costed at the list input rate (an over-estimate where caching is heavy); cached_tokens is recorded per call so a reader can recompute.",
      "latency": "observed on subscription CLI surface, indicative"
    }
  }
}