feat: add output review kit

2026-06-14 06:56:44 +08:00
parent ecdb16dd90
commit 0cf7f32e32
53 changed files with 1787 additions and 352 deletions
@@ -2,7 +2,7 @@ PYTHON ?= python3
 LOCAL_SKILL_INSTALL_DIR ?= $(HOME)/.agents/skills.disabled/yao-meta-skill
 ACTIVE_SKILL_INSTALL_DIR ?= $(HOME)/.agents/skills/yao-meta-skill

-.PHONY: eval eval-suite route-scorecard route-confusion-check description-optimization judge-blind-eval description-optimization-check promotion-check python-compat-check architecture-maintainability-check yao-cli-check skill-overview-check skill-report-metrics-check skill-report-charts-check skill-ir-check compiler-check output-eval-check output-execution-check output-review-adjudication-check runtime-conformance-check runtime-permission-check trust-check skill-atlas-check registry-audit-check package-verify-check install-simulation-check upgrade-check review-viewer-check review-studio-check skill-os2-audit-check skill-os2-coverage-check world-class-evidence-check world-class-ledger-check world-class-intake-check world-class-submission-review-check world-class-claim-guard-check benchmark-reproducibility-check feedback-check adoption-drift-check telemetry-import-check telemetry-emit-check telemetry-hooks-check telemetry-native-host-check review-waivers-check review-annotations-check baseline-compare-check reference-scan-check github-benchmark-scan-check intent-confidence-check reference-synthesis-check output-risk-profile-check artifact-design-profile-check prompt-quality-profile-check system-model-check iteration-directions-check description-drift-history iteration-ledger results-panel regression-history context-reports portability-report portability-check failure-regression-check package-check package-failure-check security-boundary-check local-install-sync-check snapshot-check validate lint governance-check resource-boundary-check quality-check sync-local-install sync-active-install test ci-test clean
+.PHONY: eval eval-suite route-scorecard route-confusion-check description-optimization judge-blind-eval description-optimization-check promotion-check python-compat-check architecture-maintainability-check yao-cli-check skill-overview-check skill-report-metrics-check skill-report-charts-check skill-ir-check compiler-check output-eval-check output-execution-check output-review-kit-check output-review-adjudication-check runtime-conformance-check runtime-permission-check trust-check skill-atlas-check registry-audit-check package-verify-check install-simulation-check upgrade-check review-viewer-check review-studio-check skill-os2-audit-check skill-os2-coverage-check world-class-evidence-check world-class-ledger-check world-class-intake-check world-class-submission-review-check world-class-claim-guard-check benchmark-reproducibility-check feedback-check adoption-drift-check telemetry-import-check telemetry-emit-check telemetry-hooks-check telemetry-native-host-check review-waivers-check review-annotations-check baseline-compare-check reference-scan-check github-benchmark-scan-check intent-confidence-check reference-synthesis-check output-risk-profile-check artifact-design-profile-check prompt-quality-profile-check system-model-check iteration-directions-check description-drift-history iteration-ledger results-panel regression-history context-reports portability-report portability-check failure-regression-check package-check package-failure-check security-boundary-check local-install-sync-check snapshot-check validate lint governance-check resource-boundary-check quality-check sync-local-install sync-active-install test ci-test clean

 eval:
 	$(PYTHON) scripts/trigger_eval.py --description-file evals/improved_description.txt --cases evals/trigger_cases.json --baseline-description-file evals/baseline_description.txt
@@ -58,6 +58,9 @@ output-eval-check:
 output-execution-check:
 	$(PYTHON) tests/verify_output_execution_runs.py

+output-review-kit-check:
+	$(PYTHON) tests/verify_output_review_kit.py
+
 output-review-adjudication-check:
 	$(PYTHON) tests/verify_output_review_adjudication.py

@@ -16,7 +16,7 @@
        "vscode"
      ],
      "package_metadata": "registry/packages/yao-meta-skill.json",
-      "package_sha256": "68a0228b56de8c9e9e94d64d7c1e386bf3dff55c0f6228b65f91e75d02525926"
+      "package_sha256": "33ca97e8df3e3c75228595ab8bd06d65c697ef708723a0a53637bd8f22bbcfda"
    }
  ]
 }
@@ -16,7 +16,7 @@
  "trust_level": "local",
  "license": "MIT",
  "checksums": {
-    "package_sha256": "68a0228b56de8c9e9e94d64d7c1e386bf3dff55c0f6228b65f91e75d02525926",
+    "package_sha256": "33ca97e8df3e3c75228595ab8bd06d65c697ef708723a0a53637bd8f22bbcfda",
    "archive_sha256": "cb1ac8c62bc5dee942a88f18eb5d71c93ae58f854ac19f3d912f9c9c3a4ec24e"
  },
  "compatibility": {
@@ -1,7 +1,7 @@
 {
  "ok": true,
  "schema_version": "2.0",
-  "generated_at": "2026-06-13T22:37:14Z",
+  "generated_at": "2026-06-13T22:55:26Z",
  "skill_dir": ".",
  "privacy_contract": {
    "storage": "local-first",
@@ -4,12 +4,12 @@
  "generated_at": "2026-06-13",
  "skill_dir": ".",
  "summary": {
-    "python_file_count": 149,
-    "script_file_count": 92,
-    "test_file_count": 57,
+    "python_file_count": 151,
+    "script_file_count": 93,
+    "test_file_count": 58,
    "internal_module_count": 19,
-    "cli_script_count": 74,
-    "command_handler_count": 33,
+    "cli_script_count": 75,
+    "command_handler_count": 34,
    "warn_line_threshold": 900,
    "block_line_threshold": 1500,
    "largest_file_lines": 898,
@@ -27,21 +27,21 @@
    },
    {
      "path": "tests/verify_yao_cli.py",
-      "lines": 891,
+      "lines": 893,
      "kind": "test",
      "severity": "pass",
      "recommendation": "Break broad integration assertions into focused verifier helpers when the next behavior change lands."
    },
    {
      "path": "scripts/yao.py",
-      "lines": 841,
+      "lines": 862,
      "kind": "cli-script",
      "severity": "pass",
      "recommendation": "Split command handlers by domain while keeping scripts/yao.py as the thin CLI orchestrator."
    },
    {
      "path": "scripts/skill_report_model.py",
-      "lines": 782,
+      "lines": 792,
      "kind": "internal-module",
      "severity": "pass",
      "recommendation": "Watch this file before adding new responsibilities; extract a helper module when one concern dominates."
@@ -53,6 +53,13 @@
      "severity": "pass",
      "recommendation": "Watch this file before adding new responsibilities; extract a helper module when one concern dominates."
    },
+    {
+      "path": "scripts/yao_cli_parser.py",
+      "lines": 726,
+      "kind": "internal-module",
+      "severity": "pass",
+      "recommendation": "Watch this file before adding new responsibilities; extract a helper module when one concern dominates."
+    },
    {
      "path": "scripts/optimize_description.py",
      "lines": 723,
@@ -67,13 +74,6 @@
      "severity": "pass",
      "recommendation": "Watch this file before adding new responsibilities; extract a helper module when one concern dominates."
    },
-    {
-      "path": "scripts/yao_cli_parser.py",
-      "lines": 714,
-      "kind": "internal-module",
-      "severity": "pass",
-      "recommendation": "Watch this file before adding new responsibilities; extract a helper module when one concern dominates."
-    },
    {
      "path": "scripts/render_review_viewer.py",
      "lines": 685,
@@ -5,12 +5,12 @@ Generated at: `2026-06-13`
 ## Summary

 - decision: `pass`
- python files: `149`
- scripts: `92`
- tests: `57`
+- python files: `151`
+- scripts: `93`
+- tests: `58`
 - internal modules: `19`
- CLI scripts: `74`
- Yao CLI command handlers: `33`
+- CLI scripts: `75`
+- Yao CLI command handlers: `34`
 - largest file lines: `898`
 - hotspots: `0`
 - blockers: `0`
@@ -26,13 +26,13 @@ No file-size hotspots found.
 | File | Lines | Kind | Severity |
 | --- | ---: | --- | --- |
 | `scripts/render_review_studio.py` | `898` | `cli-script` | `pass` |
-| `tests/verify_yao_cli.py` | `891` | `test` | `pass` |
-| `scripts/yao.py` | `841` | `cli-script` | `pass` |
-| `scripts/skill_report_model.py` | `782` | `internal-module` | `pass` |
+| `tests/verify_yao_cli.py` | `893` | `test` | `pass` |
+| `scripts/yao.py` | `862` | `cli-script` | `pass` |
+| `scripts/skill_report_model.py` | `792` | `internal-module` | `pass` |
 | `scripts/compile_skill.py` | `734` | `cli-script` | `pass` |
+| `scripts/yao_cli_parser.py` | `726` | `internal-module` | `pass` |
 | `scripts/optimize_description.py` | `723` | `cli-script` | `pass` |
 | `scripts/trust_check.py` | `714` | `internal-module` | `pass` |
-| `scripts/yao_cli_parser.py` | `714` | `internal-module` | `pass` |
 | `scripts/render_review_viewer.py` | `685` | `cli-script` | `pass` |
 | `scripts/build_skill_atlas.py` | `674` | `cli-script` | `pass` |
 | `scripts/skill_report_layout.py` | `653` | `internal-module` | `pass` |
@@ -3,16 +3,13 @@
  "ok": true,
  "generated_at": "2026-06-14",
  "skill_dir": ".",
-  "commit": "30539bed7a29dd88c2f289dd0c115da9834dd5d9",
+  "commit": "ecdb16dd90025bcca3051be4c6a1b5b723c7ec75",
  "git_status": {
    "available": true,
    "dirty": true,
-    "changed_file_count": 60,
+    "changed_file_count": 59,
    "sample": [
-      " M AGENTS.md",
      " M Makefile",
-      " M README.md",
-      " M evidence/world_class/README.md",
      " M registry/index.json",
      " M registry/packages/yao-meta-skill.json",
      " M reports/adoption_drift_report.json",
@@ -20,7 +17,10 @@
      " M reports/architecture_maintainability.json",
      " M reports/architecture_maintainability.md",
      " M reports/benchmark_reproducibility.json",
-      " M reports/benchmark_reproducibility.md"
+      " M reports/benchmark_reproducibility.md",
+      " M reports/compiled_targets.json",
+      " M reports/context_budget.json",
+      " M reports/install_simulation.json"
    ]
  },
  "summary": {
@@ -42,7 +42,7 @@
    "world_class_task_count": 4,
    "world_class_ledger_pending_count": 4,
    "working_tree_dirty": true,
-    "changed_file_count": 60
+    "changed_file_count": 59
  },
  "methodology": {
    "path": "reports/benchmark_methodology.md",
@@ -116,7 +116,7 @@
      "path": "reports/output_execution_runs.json",
      "exists": true,
      "bytes": 7967,
-      "sha256": "34fa0a084b7537cca412b4bbea4018639ceac9e98ddf83a6d25ee2dc7f7dbe4f"
+      "sha256": "b86a205e0eee68676a822c4f818b5ea0496c381c783043832be72050d5a1e716"
    },
    {
      "label": "blind_review",
@@ -129,8 +129,8 @@
      "label": "review_adjudication",
      "path": "reports/output_review_adjudication.json",
      "exists": true,
-      "bytes": 9125,
-      "sha256": "5765254c430947c7e2b05883831d7db0b886d607b59f9a8a2602dcfa925f400b"
+      "bytes": 9495,
+      "sha256": "240485a721af49d5c75fe8049e9ef74b50546f685aa6346ab9e20713fb1105f4"
    },
    {
      "label": "trigger_scorecard",
@@ -150,22 +150,22 @@
      "label": "trust_report",
      "path": "reports/security_trust_report.json",
      "exists": true,
-      "bytes": 94809,
-      "sha256": "1d256935f965a265ae5c60e8600a081db8aebe1fdeff7068c5d08fb521fe653b"
+      "bytes": 95873,
+      "sha256": "6205af0ad3c9980eab785d7644832b1ca804d898eebcf993cced83c23b256fa4"
    },
    {
      "label": "python_compatibility",
      "path": "reports/python_compatibility.json",
      "exists": true,
-      "bytes": 19840,
-      "sha256": "9b8f5b17fe2711b17557366edda49a89bbf2e378576de52d7de798137dc8a421"
+      "bytes": 20093,
+      "sha256": "e5dd973d0b5c78c5e40d52727939134b8167d742cb90bea2141744c0f7d5fa33"
    },
    {
      "label": "registry_audit",
      "path": "reports/registry_audit.json",
      "exists": true,
      "bytes": 3183,
-      "sha256": "c3ae3224b9339aec1e01e695835aa14113bf23979100eda2099fa9609e841c17"
+      "sha256": "daedc9b8047e2926221f39209d0751dac67b160c2c68fded0fcca796d9727144"
    },
    {
      "label": "package_verification",
@@ -179,42 +179,42 @@
      "path": "reports/install_simulation.json",
      "exists": true,
      "bytes": 8758,
-      "sha256": "465f407ccc808169f5380550161f481ab8f0fbc20ef9ab170814fb703b05fe3b"
+      "sha256": "3c9306b5b1221f5951f69598c8e7104557ea17635f11c42b8b52a37cae6d22a2"
    },
    {
      "label": "skill_os2_audit",
      "path": "reports/skill_os2_audit.json",
      "exists": true,
      "bytes": 13946,
-      "sha256": "340c10df7350d74cbfe3fd77461ce2cbab201198fd93fd276e7273eb6f6222e7"
+      "sha256": "9c22f67024a63d8646c2306ef642e06848bb443adde690e2f97573024e3a636b"
    },
    {
      "label": "world_class_evidence_plan",
      "path": "reports/world_class_evidence_plan.json",
      "exists": true,
-      "bytes": 9532,
-      "sha256": "ed1274b7c18cac9e683bb4a24a5ed8039d1eb582d676f0fc26cc3a33e5371a67"
+      "bytes": 9634,
+      "sha256": "7ff33ae06f00247f97fa85ca3268aaa61f672aebba75ac96fc3d932b4997daa2"
    },
    {
      "label": "world_class_evidence_ledger",
      "path": "reports/world_class_evidence_ledger.json",
      "exists": true,
-      "bytes": 10750,
-      "sha256": "e6dfdb22dc9f54255c1d4b42d36c035ffafbca4fe1d1bc6569c5e8262443f257"
+      "bytes": 10790,
+      "sha256": "2e2dc29095a05d481353fec00f5b87c3381c81ba9661fb36b03496d5e5db325e"
    },
    {
      "label": "world_class_evidence_intake",
      "path": "reports/world_class_evidence_intake.json",
      "exists": true,
-      "bytes": 12369,
-      "sha256": "465f8d915a869ad536c2d71ba26dc536b82ba64e6e98013e8dbd0efd1bb92883"
+      "bytes": 12411,
+      "sha256": "10ac53e66486505ff1646f5616ffa334c0e2a2ac99d9d54cf6222d2ee3868063"
    },
    {
      "label": "world_class_claim_guard",
      "path": "reports/world_class_claim_guard.json",
      "exists": true,
-      "bytes": 8093,
-      "sha256": "24396b9ecb4351cb9e3dff3d69b83bb940bf1507387a086cb0a08ca8cfdfa618"
+      "bytes": 8179,
+      "sha256": "299a88eb7b47d16761d66980e01ccd8164609e344c3c98bb0b8a67b56cf68edd"
    }
  ],
  "missing_artifacts": [],
@@ -1,7 +1,7 @@
 # Benchmark Reproducibility

 Generated at: `2026-06-14`
-Commit: `30539bed7a29dd88c2f289dd0c115da9834dd5d9`
+Commit: `ecdb16dd90025bcca3051be4c6a1b5b723c7ec75`
 Working tree dirty at generation: `true`

 ## Summary
@@ -16,7 +16,7 @@ Working tree dirty at generation: `true`
 - provider evidence complete: `false`
 - human review complete: `false`
 - world-class ready: `false`
- changed files at generation: `60`
+- changed files at generation: `59`

 This report proves local benchmark reproducibility only. It keeps external provider and human-review gaps visible instead of counting them as complete.

@@ -40,21 +40,21 @@ This report proves local benchmark reproducibility only. It keeps external provi
 | output_cases | `evals/output/cases.jsonl` | present | `a6ae96857116` |
 | output_schema | `evals/output/schema.json` | present | `8ee340c95064` |
 | output_scorecard | `reports/output_quality_scorecard.json` | present | `0806258a8e08` |
-| output_execution | `reports/output_execution_runs.json` | present | `34fa0a084b75` |
+| output_execution | `reports/output_execution_runs.json` | present | `b86a205e0eee` |
 | blind_review | `reports/output_blind_review_pack.json` | present | `bbe2db8ec277` |
-| review_adjudication | `reports/output_review_adjudication.json` | present | `5765254c4309` |
+| review_adjudication | `reports/output_review_adjudication.json` | present | `240485a721af` |
 | trigger_scorecard | `reports/route_scorecard.json` | present | `c164e83e36d0` |
 | runtime_conformance | `reports/conformance_matrix.json` | present | `8251329e663d` |
-| trust_report | `reports/security_trust_report.json` | present | `1d256935f965` |
-| python_compatibility | `reports/python_compatibility.json` | present | `9b8f5b17fe27` |
-| registry_audit | `reports/registry_audit.json` | present | `c3ae3224b933` |
+| trust_report | `reports/security_trust_report.json` | present | `6205af0ad3c9` |
+| python_compatibility | `reports/python_compatibility.json` | present | `e5dd973d0b5c` |
+| registry_audit | `reports/registry_audit.json` | present | `daedc9b8047e` |
 | package_verification | `reports/package_verification.json` | present | `89dbb682f4a4` |
-| install_simulation | `reports/install_simulation.json` | present | `465f407ccc80` |
-| skill_os2_audit | `reports/skill_os2_audit.json` | present | `340c10df7350` |
-| world_class_evidence_plan | `reports/world_class_evidence_plan.json` | present | `ed1274b7c18c` |
-| world_class_evidence_ledger | `reports/world_class_evidence_ledger.json` | present | `e6dfdb22dc9f` |
-| world_class_evidence_intake | `reports/world_class_evidence_intake.json` | present | `465f8d915a86` |
-| world_class_claim_guard | `reports/world_class_claim_guard.json` | present | `24396b9ecb43` |
+| install_simulation | `reports/install_simulation.json` | present | `3c9306b5b122` |
+| skill_os2_audit | `reports/skill_os2_audit.json` | present | `9c22f67024a6` |
+| world_class_evidence_plan | `reports/world_class_evidence_plan.json` | present | `7ff33ae06f00` |
+| world_class_evidence_ledger | `reports/world_class_evidence_ledger.json` | present | `2e2dc29095a0` |
+| world_class_evidence_intake | `reports/world_class_evidence_intake.json` | present | `10ac53e66486` |
+| world_class_claim_guard | `reports/world_class_claim_guard.json` | present | `299a88eb7b47` |

 ## Reproduction Commands

@@ -166,6 +166,7 @@
            "scripts/lint_skill.py",
            "scripts/local_output_eval_runner.py",
            "scripts/optimize_description.py",
+            "scripts/prepare_output_review_kit.py",
            "scripts/prepare_world_class_submission_kit.py",
            "scripts/probe_runtime_permissions.py",
            "scripts/promotion_checker.py",
@@ -223,8 +224,7 @@
            "scripts/skill_report_layout.py",
            "scripts/skill_report_metrics.py",
            "scripts/skill_report_model.py",
-            "scripts/sync_local_install.py",
-            "scripts/telemetry_native_host.py"
+            "scripts/sync_local_install.py"
          ],
          "assets": [
            "templates/basic_skill.md.j2",
@@ -383,7 +383,7 @@
            },
            "file_write": {
              "required": true,
-              "script_count": 62,
+              "script_count": 63,
              "scripts": [
                "scripts/adjudicate_output_review.py",
                "scripts/build_confusion_matrix.py",
@@ -399,6 +399,7 @@
                "scripts/github_benchmark_scan.py",
                "scripts/import_telemetry_events.py",
                "scripts/init_skill.py",
+                "scripts/prepare_output_review_kit.py",
                "scripts/prepare_world_class_submission_kit.py",
                "scripts/probe_runtime_permissions.py",
                "scripts/promotion_checker.py",
@@ -485,14 +486,14 @@
          },
          "help_smoke": {
            "enabled": true,
-            "checked_count": 75,
+            "checked_count": 76,
            "failed_count": 0,
            "failed_scripts": []
          },
          "trust_summary": {
            "secret_findings": 0,
            "network_script_count": 3,
-            "file_write_script_count": 62,
+            "file_write_script_count": 63,
            "subprocess_script_count": 9,
            "interactive_script_count": 0,
            "help_smoke_failed_count": 0
@@ -514,7 +515,7 @@
          ],
          "capability_counts": {
            "network": 3,
-            "file_write": 62,
+            "file_write": 63,
            "subprocess": 9,
            "interactive": 0
          },
@@ -627,7 +628,7 @@
          },
          "file_write": {
            "required": true,
-            "script_count": 62,
+            "script_count": 63,
            "scripts": [
              "scripts/adjudicate_output_review.py",
              "scripts/build_confusion_matrix.py",
@@ -643,6 +644,7 @@
              "scripts/github_benchmark_scan.py",
              "scripts/import_telemetry_events.py",
              "scripts/init_skill.py",
+              "scripts/prepare_output_review_kit.py",
              "scripts/prepare_world_class_submission_kit.py",
              "scripts/probe_runtime_permissions.py",
              "scripts/promotion_checker.py",
@@ -729,14 +731,14 @@
        },
        "help_smoke": {
          "enabled": true,
-          "checked_count": 75,
+          "checked_count": 76,
          "failed_count": 0,
          "failed_scripts": []
        },
        "trust_summary": {
          "secret_findings": 0,
          "network_script_count": 3,
-          "file_write_script_count": 62,
+          "file_write_script_count": 63,
          "subprocess_script_count": 9,
          "interactive_script_count": 0,
          "help_smoke_failed_count": 0
@@ -756,7 +758,7 @@
        ],
        "capability_counts": {
          "network": 3,
-          "file_write": 62,
+          "file_write": 63,
          "subprocess": 9,
          "interactive": 0
        },
@@ -1011,6 +1013,7 @@
            "scripts/lint_skill.py",
            "scripts/local_output_eval_runner.py",
            "scripts/optimize_description.py",
+            "scripts/prepare_output_review_kit.py",
            "scripts/prepare_world_class_submission_kit.py",
            "scripts/probe_runtime_permissions.py",
            "scripts/promotion_checker.py",
@@ -1068,8 +1071,7 @@
            "scripts/skill_report_layout.py",
            "scripts/skill_report_metrics.py",
            "scripts/skill_report_model.py",
-            "scripts/sync_local_install.py",
-            "scripts/telemetry_native_host.py"
+            "scripts/sync_local_install.py"
          ],
          "assets": [
            "templates/basic_skill.md.j2",
@@ -1228,7 +1230,7 @@
            },
            "file_write": {
              "required": true,
-              "script_count": 62,
+              "script_count": 63,
              "scripts": [
                "scripts/adjudicate_output_review.py",
                "scripts/build_confusion_matrix.py",
@@ -1244,6 +1246,7 @@
                "scripts/github_benchmark_scan.py",
                "scripts/import_telemetry_events.py",
                "scripts/init_skill.py",
+                "scripts/prepare_output_review_kit.py",
                "scripts/prepare_world_class_submission_kit.py",
                "scripts/probe_runtime_permissions.py",
                "scripts/promotion_checker.py",
@@ -1330,14 +1333,14 @@
          },
          "help_smoke": {
            "enabled": true,
-            "checked_count": 75,
+            "checked_count": 76,
            "failed_count": 0,
            "failed_scripts": []
          },
          "trust_summary": {
            "secret_findings": 0,
            "network_script_count": 3,
-            "file_write_script_count": 62,
+            "file_write_script_count": 63,
            "subprocess_script_count": 9,
            "interactive_script_count": 0,
            "help_smoke_failed_count": 0
@@ -1359,7 +1362,7 @@
          ],
          "capability_counts": {
            "network": 3,
-            "file_write": 62,
+            "file_write": 63,
            "subprocess": 9,
            "interactive": 0
          },
@@ -1472,7 +1475,7 @@
          },
          "file_write": {
            "required": true,
-            "script_count": 62,
+            "script_count": 63,
            "scripts": [
              "scripts/adjudicate_output_review.py",
              "scripts/build_confusion_matrix.py",
@@ -1488,6 +1491,7 @@
              "scripts/github_benchmark_scan.py",
              "scripts/import_telemetry_events.py",
              "scripts/init_skill.py",
+              "scripts/prepare_output_review_kit.py",
              "scripts/prepare_world_class_submission_kit.py",
              "scripts/probe_runtime_permissions.py",
              "scripts/promotion_checker.py",
@@ -1574,14 +1578,14 @@
        },
        "help_smoke": {
          "enabled": true,
-          "checked_count": 75,
+          "checked_count": 76,
          "failed_count": 0,
          "failed_scripts": []
        },
        "trust_summary": {
          "secret_findings": 0,
          "network_script_count": 3,
-          "file_write_script_count": 62,
+          "file_write_script_count": 63,
          "subprocess_script_count": 9,
          "interactive_script_count": 0,
          "help_smoke_failed_count": 0
@@ -1601,7 +1605,7 @@
        ],
        "capability_counts": {
          "network": 3,
-          "file_write": 62,
+          "file_write": 63,
          "subprocess": 9,
          "interactive": 0
        },
@@ -1856,6 +1860,7 @@
            "scripts/lint_skill.py",
            "scripts/local_output_eval_runner.py",
            "scripts/optimize_description.py",
+            "scripts/prepare_output_review_kit.py",
            "scripts/prepare_world_class_submission_kit.py",
            "scripts/probe_runtime_permissions.py",
            "scripts/promotion_checker.py",
@@ -1913,8 +1918,7 @@
            "scripts/skill_report_layout.py",
            "scripts/skill_report_metrics.py",
            "scripts/skill_report_model.py",
-            "scripts/sync_local_install.py",
-            "scripts/telemetry_native_host.py"
+            "scripts/sync_local_install.py"
          ],
          "assets": [
            "templates/basic_skill.md.j2",
@@ -2073,7 +2077,7 @@
            },
            "file_write": {
              "required": true,
-              "script_count": 62,
+              "script_count": 63,
              "scripts": [
                "scripts/adjudicate_output_review.py",
                "scripts/build_confusion_matrix.py",
@@ -2089,6 +2093,7 @@
                "scripts/github_benchmark_scan.py",
                "scripts/import_telemetry_events.py",
                "scripts/init_skill.py",
+                "scripts/prepare_output_review_kit.py",
                "scripts/prepare_world_class_submission_kit.py",
                "scripts/probe_runtime_permissions.py",
                "scripts/promotion_checker.py",
@@ -2175,14 +2180,14 @@
          },
          "help_smoke": {
            "enabled": true,
-            "checked_count": 75,
+            "checked_count": 76,
            "failed_count": 0,
            "failed_scripts": []
          },
          "trust_summary": {
            "secret_findings": 0,
            "network_script_count": 3,
-            "file_write_script_count": 62,
+            "file_write_script_count": 63,
            "subprocess_script_count": 9,
            "interactive_script_count": 0,
            "help_smoke_failed_count": 0
@@ -2204,7 +2209,7 @@
          ],
          "capability_counts": {
            "network": 3,
-            "file_write": 62,
+            "file_write": 63,
            "subprocess": 9,
            "interactive": 0
          },
@@ -2310,7 +2315,7 @@
          },
          "file_write": {
            "required": true,
-            "script_count": 62,
+            "script_count": 63,
            "scripts": [
              "scripts/adjudicate_output_review.py",
              "scripts/build_confusion_matrix.py",
@@ -2326,6 +2331,7 @@
              "scripts/github_benchmark_scan.py",
              "scripts/import_telemetry_events.py",
              "scripts/init_skill.py",
+              "scripts/prepare_output_review_kit.py",
              "scripts/prepare_world_class_submission_kit.py",
              "scripts/probe_runtime_permissions.py",
              "scripts/promotion_checker.py",
@@ -2412,14 +2418,14 @@
        },
        "help_smoke": {
          "enabled": true,
-          "checked_count": 75,
+          "checked_count": 76,
          "failed_count": 0,
          "failed_scripts": []
        },
        "trust_summary": {
          "secret_findings": 0,
          "network_script_count": 3,
-          "file_write_script_count": 62,
+          "file_write_script_count": 63,
          "subprocess_script_count": 9,
          "interactive_script_count": 0,
          "help_smoke_failed_count": 0
@@ -2439,7 +2445,7 @@
        ],
        "capability_counts": {
          "network": 3,
-          "file_write": 62,
+          "file_write": 63,
          "subprocess": 9,
          "interactive": 0
        },
@@ -2685,6 +2691,7 @@
            "scripts/lint_skill.py",
            "scripts/local_output_eval_runner.py",
            "scripts/optimize_description.py",
+            "scripts/prepare_output_review_kit.py",
            "scripts/prepare_world_class_submission_kit.py",
            "scripts/probe_runtime_permissions.py",
            "scripts/promotion_checker.py",
@@ -2742,8 +2749,7 @@
            "scripts/skill_report_layout.py",
            "scripts/skill_report_metrics.py",
            "scripts/skill_report_model.py",
-            "scripts/sync_local_install.py",
-            "scripts/telemetry_native_host.py"
+            "scripts/sync_local_install.py"
          ],
          "assets": [
            "templates/basic_skill.md.j2",
@@ -2902,7 +2908,7 @@
            },
            "file_write": {
              "required": true,
-              "script_count": 62,
+              "script_count": 63,
              "scripts": [
                "scripts/adjudicate_output_review.py",
                "scripts/build_confusion_matrix.py",
@@ -2918,6 +2924,7 @@
                "scripts/github_benchmark_scan.py",
                "scripts/import_telemetry_events.py",
                "scripts/init_skill.py",
+                "scripts/prepare_output_review_kit.py",
                "scripts/prepare_world_class_submission_kit.py",
                "scripts/probe_runtime_permissions.py",
                "scripts/promotion_checker.py",
@@ -3004,14 +3011,14 @@
          },
          "help_smoke": {
            "enabled": true,
-            "checked_count": 75,
+            "checked_count": 76,
            "failed_count": 0,
            "failed_scripts": []
          },
          "trust_summary": {
            "secret_findings": 0,
            "network_script_count": 3,
-            "file_write_script_count": 62,
+            "file_write_script_count": 63,
            "subprocess_script_count": 9,
            "interactive_script_count": 0,
            "help_smoke_failed_count": 0
@@ -3033,7 +3040,7 @@
          ],
          "capability_counts": {
            "network": 3,
-            "file_write": 62,
+            "file_write": 63,
            "subprocess": 9,
            "interactive": 0
          },
@@ -3139,7 +3146,7 @@
          },
          "file_write": {
            "required": true,
-            "script_count": 62,
+            "script_count": 63,
            "scripts": [
              "scripts/adjudicate_output_review.py",
              "scripts/build_confusion_matrix.py",
@@ -3155,6 +3162,7 @@
              "scripts/github_benchmark_scan.py",
              "scripts/import_telemetry_events.py",
              "scripts/init_skill.py",
+              "scripts/prepare_output_review_kit.py",
              "scripts/prepare_world_class_submission_kit.py",
              "scripts/probe_runtime_permissions.py",
              "scripts/promotion_checker.py",
@@ -3241,14 +3249,14 @@
        },
        "help_smoke": {
          "enabled": true,
-          "checked_count": 75,
+          "checked_count": 76,
          "failed_count": 0,
          "failed_scripts": []
        },
        "trust_summary": {
          "secret_findings": 0,
          "network_script_count": 3,
-          "file_write_script_count": 62,
+          "file_write_script_count": 63,
          "subprocess_script_count": 9,
          "interactive_script_count": 0,
          "help_smoke_failed_count": 0
@@ -3268,7 +3276,7 @@
        ],
        "capability_counts": {
          "network": 3,
-          "file_write": 62,
+          "file_write": 63,
          "subprocess": 9,
          "interactive": 0
        },
@@ -3514,6 +3522,7 @@
            "scripts/lint_skill.py",
            "scripts/local_output_eval_runner.py",
            "scripts/optimize_description.py",
+            "scripts/prepare_output_review_kit.py",
            "scripts/prepare_world_class_submission_kit.py",
            "scripts/probe_runtime_permissions.py",
            "scripts/promotion_checker.py",
@@ -3571,8 +3580,7 @@
            "scripts/skill_report_layout.py",
            "scripts/skill_report_metrics.py",
            "scripts/skill_report_model.py",
-            "scripts/sync_local_install.py",
-            "scripts/telemetry_native_host.py"
+            "scripts/sync_local_install.py"
          ],
          "assets": [
            "templates/basic_skill.md.j2",
@@ -3731,7 +3739,7 @@
            },
            "file_write": {
              "required": true,
-              "script_count": 62,
+              "script_count": 63,
              "scripts": [
                "scripts/adjudicate_output_review.py",
                "scripts/build_confusion_matrix.py",
@@ -3747,6 +3755,7 @@
                "scripts/github_benchmark_scan.py",
                "scripts/import_telemetry_events.py",
                "scripts/init_skill.py",
+                "scripts/prepare_output_review_kit.py",
                "scripts/prepare_world_class_submission_kit.py",
                "scripts/probe_runtime_permissions.py",
                "scripts/promotion_checker.py",
@@ -3833,14 +3842,14 @@
          },
          "help_smoke": {
            "enabled": true,
-            "checked_count": 75,
+            "checked_count": 76,
            "failed_count": 0,
            "failed_scripts": []
          },
          "trust_summary": {
            "secret_findings": 0,
            "network_script_count": 3,
-            "file_write_script_count": 62,
+            "file_write_script_count": 63,
            "subprocess_script_count": 9,
            "interactive_script_count": 0,
            "help_smoke_failed_count": 0
@@ -3862,7 +3871,7 @@
          ],
          "capability_counts": {
            "network": 3,
-            "file_write": 62,
+            "file_write": 63,
            "subprocess": 9,
            "interactive": 0
          },
@@ -3972,7 +3981,7 @@
          },
          "file_write": {
            "required": true,
-            "script_count": 62,
+            "script_count": 63,
            "scripts": [
              "scripts/adjudicate_output_review.py",
              "scripts/build_confusion_matrix.py",
@@ -3988,6 +3997,7 @@
              "scripts/github_benchmark_scan.py",
              "scripts/import_telemetry_events.py",
              "scripts/init_skill.py",
+              "scripts/prepare_output_review_kit.py",
              "scripts/prepare_world_class_submission_kit.py",
              "scripts/probe_runtime_permissions.py",
              "scripts/promotion_checker.py",
@@ -4074,14 +4084,14 @@
        },
        "help_smoke": {
          "enabled": true,
-          "checked_count": 75,
+          "checked_count": 76,
          "failed_count": 0,
          "failed_scripts": []
        },
        "trust_summary": {
          "secret_findings": 0,
          "network_script_count": 3,
-          "file_write_script_count": 62,
+          "file_write_script_count": 63,
          "subprocess_script_count": 9,
          "interactive_script_count": 0,
          "help_smoke_failed_count": 0
@@ -4101,7 +4111,7 @@
        ],
        "capability_counts": {
          "network": 3,
-          "file_write": 62,
+          "file_write": 63,
          "subprocess": 9,
          "interactive": 0
        },
@@ -6,10 +6,10 @@
    "context_budget_tier": "production",
    "context_budget_limit": 1000,
    "skill_body_tokens": 751,
-    "other_text_tokens": 1129811,
+    "other_text_tokens": 1145183,
    "estimated_initial_load_tokens": 944,
-    "estimated_total_text_tokens": 1130562,
-    "relevant_file_count": 484,
+    "estimated_total_text_tokens": 1145934,
+    "relevant_file_count": 490,
    "unused_resource_dirs": [],
    "quality_signal_points": 130,
    "quality_density": 137.7
@@ -8,7 +8,7 @@
  "installed_skill_dir": "tests/tmp_review_studio/install-root/simulate-yao-meta-skill/yao-meta-skill",
  "summary": {
    "archive_present": true,
-    "archive_entry_count": 563,
+    "archive_entry_count": 567,
    "archive_extracted": true,
    "entrypoint_loaded": true,
    "manifest_loaded": true,
@@ -34,7 +34,7 @@
      "execution_mode": "command",
      "model_executed": false,
      "command_executed": true,
-      "duration_ms": 27.77,
+      "duration_ms": 27.61,
      "provider": "local-output-eval-runner",
      "model": "",
      "usage": {
@@ -62,7 +62,7 @@
      "execution_mode": "command",
      "model_executed": false,
      "command_executed": true,
-      "duration_ms": 28.86,
+      "duration_ms": 26.83,
      "provider": "local-output-eval-runner",
      "model": "",
      "usage": {
@@ -85,7 +85,7 @@
      "execution_mode": "command",
      "model_executed": false,
      "command_executed": true,
-      "duration_ms": 28.67,
+      "duration_ms": 28.0,
      "provider": "local-output-eval-runner",
      "model": "",
      "usage": {
@@ -113,7 +113,7 @@
      "execution_mode": "command",
      "model_executed": false,
      "command_executed": true,
-      "duration_ms": 28.37,
+      "duration_ms": 28.5,
      "provider": "local-output-eval-runner",
      "model": "",
      "usage": {
@@ -136,7 +136,7 @@
      "execution_mode": "command",
      "model_executed": false,
      "command_executed": true,
-      "duration_ms": 28.7,
+      "duration_ms": 28.58,
      "provider": "local-output-eval-runner",
      "model": "",
      "usage": {
@@ -164,7 +164,7 @@
      "execution_mode": "command",
      "model_executed": false,
      "command_executed": true,
-      "duration_ms": 27.96,
+      "duration_ms": 28.82,
      "provider": "local-output-eval-runner",
      "model": "",
      "usage": {
@@ -187,7 +187,7 @@
      "execution_mode": "command",
      "model_executed": false,
      "command_executed": true,
-      "duration_ms": 28.07,
+      "duration_ms": 28.41,
      "provider": "local-output-eval-runner",
      "model": "",
      "usage": {
@@ -214,7 +214,7 @@
      "execution_mode": "command",
      "model_executed": false,
      "command_executed": true,
-      "duration_ms": 28.42,
+      "duration_ms": 28.26,
      "provider": "local-output-eval-runner",
      "model": "",
      "usage": {
@@ -237,7 +237,7 @@
      "execution_mode": "command",
      "model_executed": false,
      "command_executed": true,
-      "duration_ms": 28.64,
+      "duration_ms": 28.37,
      "provider": "local-output-eval-runner",
      "model": "",
      "usage": {
@@ -266,7 +266,7 @@
      "execution_mode": "command",
      "model_executed": false,
      "command_executed": true,
-      "duration_ms": 28.43,
+      "duration_ms": 29.97,
      "provider": "local-output-eval-runner",
      "model": "",
      "usage": {
@@ -23,16 +23,16 @@ Command runner evidence is present. This proves the eval harness executed an ext

 | Case | Variant | Mode | Model | Duration ms | Tokens | Score | Status |
 | --- | --- | --- | --- | ---: | ---: | ---: | --- |
-| skill-package-contract | baseline | command | local-output-eval-runner | 27.77 | 33 | 0.0 | pass |
-| skill-package-contract | with_skill | command | local-output-eval-runner | 28.86 | 73 | 100.0 | pass |
-| output-eval-expectation | baseline | command | local-output-eval-runner | 28.67 | 36 | 0.0 | pass |
-| output-eval-expectation | with_skill | command | local-output-eval-runner | 28.37 | 80 | 100.0 | pass |
-| ir-before-packaging | baseline | command | local-output-eval-runner | 28.7 | 33 | 0.0 | pass |
-| ir-before-packaging | with_skill | command | local-output-eval-runner | 27.96 | 80 | 100.0 | pass |
-| near-neighbor-boundary | baseline | command | local-output-eval-runner | 28.07 | 36 | 0.0 | pass |
-| near-neighbor-boundary | with_skill | command | local-output-eval-runner | 28.42 | 65 | 100.0 | pass |
-| file-backed-governed-package | baseline | command | local-output-eval-runner | 28.64 | 37 | 0.0 | pass |
-| file-backed-governed-package | with_skill | command | local-output-eval-runner | 28.43 | 98 | 100.0 | pass |
+| skill-package-contract | baseline | command | local-output-eval-runner | 27.61 | 33 | 0.0 | pass |
+| skill-package-contract | with_skill | command | local-output-eval-runner | 26.83 | 73 | 100.0 | pass |
+| output-eval-expectation | baseline | command | local-output-eval-runner | 28.0 | 36 | 0.0 | pass |
+| output-eval-expectation | with_skill | command | local-output-eval-runner | 28.5 | 80 | 100.0 | pass |
+| ir-before-packaging | baseline | command | local-output-eval-runner | 28.58 | 33 | 0.0 | pass |
+| ir-before-packaging | with_skill | command | local-output-eval-runner | 28.82 | 80 | 100.0 | pass |
+| near-neighbor-boundary | baseline | command | local-output-eval-runner | 28.41 | 36 | 0.0 | pass |
+| near-neighbor-boundary | with_skill | command | local-output-eval-runner | 28.26 | 65 | 100.0 | pass |
+| file-backed-governed-package | baseline | command | local-output-eval-runner | 28.37 | 37 | 0.0 | pass |
+| file-backed-governed-package | with_skill | command | local-output-eval-runner | 29.97 | 98 | 100.0 | pass |

 ## Next Fixes

@@ -92,6 +92,7 @@
      "blind_pack_path": "reports/output_blind_review_pack.json",
      "decisions_path": "reports/output_review_decisions.json",
      "commands": {
+        "prepare_review_kit": "python3 scripts/yao.py output-review-kit",
        "write_template": "python3 scripts/adjudicate_output_review.py --write-template",
        "adjudicate": "python3 scripts/yao.py output-review",
        "refresh_review_studio": "python3 scripts/yao.py review-studio ."
@@ -118,6 +119,7 @@
      "blind_pack_path": "reports/output_blind_review_pack.json",
      "decisions_path": "reports/output_review_decisions.json",
      "commands": {
+        "prepare_review_kit": "python3 scripts/yao.py output-review-kit",
        "write_template": "python3 scripts/adjudicate_output_review.py --write-template",
        "adjudicate": "python3 scripts/yao.py output-review",
        "refresh_review_studio": "python3 scripts/yao.py review-studio ."
@@ -144,6 +146,7 @@
      "blind_pack_path": "reports/output_blind_review_pack.json",
      "decisions_path": "reports/output_review_decisions.json",
      "commands": {
+        "prepare_review_kit": "python3 scripts/yao.py output-review-kit",
        "write_template": "python3 scripts/adjudicate_output_review.py --write-template",
        "adjudicate": "python3 scripts/yao.py output-review",
        "refresh_review_studio": "python3 scripts/yao.py review-studio ."
@@ -170,6 +173,7 @@
      "blind_pack_path": "reports/output_blind_review_pack.json",
      "decisions_path": "reports/output_review_decisions.json",
      "commands": {
+        "prepare_review_kit": "python3 scripts/yao.py output-review-kit",
        "write_template": "python3 scripts/adjudicate_output_review.py --write-template",
        "adjudicate": "python3 scripts/yao.py output-review",
        "refresh_review_studio": "python3 scripts/yao.py review-studio ."
@@ -196,6 +200,7 @@
      "blind_pack_path": "reports/output_blind_review_pack.json",
      "decisions_path": "reports/output_review_decisions.json",
      "commands": {
+        "prepare_review_kit": "python3 scripts/yao.py output-review-kit",
        "write_template": "python3 scripts/adjudicate_output_review.py --write-template",
        "adjudicate": "python3 scripts/yao.py output-review",
        "refresh_review_studio": "python3 scripts/yao.py review-studio ."
@@ -46,6 +46,7 @@ Expected winners stay hidden until a valid reviewer decision is recorded.

 #### Commands

+- prepare_review_kit: `python3 scripts/yao.py output-review-kit`
 - write_template: `python3 scripts/adjudicate_output_review.py --write-template`
 - adjudicate: `python3 scripts/yao.py output-review`
 - refresh_review_studio: `python3 scripts/yao.py review-studio .`
@@ -72,6 +73,7 @@ Expected winners stay hidden until a valid reviewer decision is recorded.

 #### Commands

+- prepare_review_kit: `python3 scripts/yao.py output-review-kit`
 - write_template: `python3 scripts/adjudicate_output_review.py --write-template`
 - adjudicate: `python3 scripts/yao.py output-review`
 - refresh_review_studio: `python3 scripts/yao.py review-studio .`
@@ -98,6 +100,7 @@ Expected winners stay hidden until a valid reviewer decision is recorded.

 #### Commands

+- prepare_review_kit: `python3 scripts/yao.py output-review-kit`
 - write_template: `python3 scripts/adjudicate_output_review.py --write-template`
 - adjudicate: `python3 scripts/yao.py output-review`
 - refresh_review_studio: `python3 scripts/yao.py review-studio .`
@@ -124,6 +127,7 @@ Expected winners stay hidden until a valid reviewer decision is recorded.

 #### Commands

+- prepare_review_kit: `python3 scripts/yao.py output-review-kit`
 - write_template: `python3 scripts/adjudicate_output_review.py --write-template`
 - adjudicate: `python3 scripts/yao.py output-review`
 - refresh_review_studio: `python3 scripts/yao.py review-studio .`
@@ -150,6 +154,7 @@ Expected winners stay hidden until a valid reviewer decision is recorded.

 #### Commands

+- prepare_review_kit: `python3 scripts/yao.py output-review-kit`
 - write_template: `python3 scripts/adjudicate_output_review.py --write-template`
 - adjudicate: `python3 scripts/yao.py output-review`
 - refresh_review_studio: `python3 scripts/yao.py review-studio .`
@@ -0,0 +1,267 @@
+{
+  "schema_version": "1.0",
+  "ok": true,
+  "summary": {
+    "case_count": 5,
+    "ready_for_adjudication_count": 0,
+    "pending_decision_count": 5,
+    "invalid_decision_count": 0,
+    "reviewer_metadata_present": false,
+    "answer_key_hidden": true,
+    "answer_key_path_exposed": false,
+    "ready_to_run_adjudication": false,
+    "failure_count": 0
+  },
+  "artifacts": {
+    "reviewer_kit_json": "reports/output_review_kit.json",
+    "reviewer_kit_markdown": "reports/output_review_kit.md",
+    "blind_pack_json": "reports/output_blind_review_pack.json",
+    "blind_pack_markdown": "reports/output_blind_review_pack.md",
+    "decisions": "reports/output_review_decisions.json",
+    "adjudication_json": "reports/output_review_adjudication.json",
+    "adjudication_markdown": "reports/output_review_adjudication.md"
+  },
+  "template_written": false,
+  "review_contract": {
+    "reviewer_steps": [
+      "Open reports/output_blind_review_pack.md or this kit and compare Variant A vs Variant B for each case.",
+      "Record choices in reports/output_review_decisions.json without opening the answer key.",
+      "Use winner_variant A or B, confidence from 0 to 1, and a short reason for every case.",
+      "Run python3 scripts/yao.py output-review after choices are recorded.",
+      "Refresh python3 scripts/yao.py review-studio . before asking for release approval."
+    ],
+    "required_fields": {
+      "reviewer": "Human reviewer name or review group.",
+      "reviewed_at": "Review date or timestamp.",
+      "winner_variant": "A or B for every case.",
+      "confidence": "Optional numeric confidence from 0 to 1.",
+      "reason": "Short rationale based on the rubric, not on hidden labels."
+    },
+    "privacy_contract": [
+      "The answer key is intentionally withheld from this kit.",
+      "Do not inspect hidden labels or expected winners before decisions are recorded.",
+      "Do not paste private user data into decision reasons.",
+      "Pending decisions must stay pending instead of being counted as human agreement."
+    ]
+  },
+  "cases": [
+    {
+      "case_id": "skill-package-contract",
+      "prompt": "Turn this repeated workflow into a reusable team skill package.",
+      "input_files": [],
+      "review_instruction": "Pick A or B based only on the rubric. Do not infer which output came from the skill.",
+      "rubric": [
+        {
+          "id": "has-entrypoint",
+          "description": "Output names the routeable Skill entrypoint.",
+          "weight": 1.0
+        },
+        {
+          "id": "has-interface",
+          "description": "Output includes neutral interface metadata.",
+          "weight": 1.0
+        },
+        {
+          "id": "has-report",
+          "description": "Output points reviewer to generated overview report.",
+          "weight": 1.0
+        },
+        {
+          "id": "has-resource-boundary",
+          "description": "Output preserves entrypoint/resource separation.",
+          "weight": 1.0
+        }
+      ],
+      "variant_a": {
+        "blind_id": "skill-package-contract:A",
+        "output": "Create a routeable SKILL.md, agents/interface.yaml, reports/skill-overview.html, and a focused eval plan. Keep the root entrypoint lean, move durable guidance into references/, scripts into scripts/, and evidence into reports/."
+      },
+      "variant_b": {
+        "blind_id": "skill-package-contract:B",
+        "output": "I can write a prompt for that workflow and include a short checklist."
+      },
+      "decision_state": {
+        "status": "awaiting-decision",
+        "winner_variant_recorded": false,
+        "confidence_recorded": false,
+        "reason_recorded": false,
+        "blocking_reason": "Decision template exists but this row is still blank."
+      }
+    },
+    {
+      "case_id": "output-eval-expectation",
+      "prompt": "Upgrade this production skill so we know whether the generated output is better than baseline.",
+      "input_files": [],
+      "review_instruction": "Pick A or B based only on the rubric. Do not infer which output came from the skill.",
+      "rubric": [
+        {
+          "id": "has-baseline-comparison",
+          "description": "Output explicitly compares with-skill and baseline outputs.",
+          "weight": 1.0
+        },
+        {
+          "id": "has-assertion-grading",
+          "description": "Output includes objective assertion grading.",
+          "weight": 1.0
+        },
+        {
+          "id": "has-scorecard",
+          "description": "Output produces a scorecard report path.",
+          "weight": 1.0
+        },
+        {
+          "id": "has-failure-taxonomy",
+          "description": "Output records failure taxonomy.",
+          "weight": 1.0
+        }
+      ],
+      "variant_a": {
+        "blind_id": "output-eval-expectation:A",
+        "output": "Add Output Eval Lab cases with baseline_output and with_skill_output, run assertion grading, report with-skill vs baseline pass-rate delta, and record failure taxonomy plus next fixes in reports/output_quality_scorecard.md."
+      },
+      "variant_b": {
+        "blind_id": "output-eval-expectation:B",
+        "output": "Add more examples and run the trigger eval again."
+      },
+      "decision_state": {
+        "status": "awaiting-decision",
+        "winner_variant_recorded": false,
+        "confidence_recorded": false,
+        "reason_recorded": false,
+        "blocking_reason": "Decision template exists but this row is still blank."
+      }
+    },
+    {
+      "case_id": "ir-before-packaging",
+      "prompt": "Package this skill for OpenAI, Claude, Agent Skills, and generic targets.",
+      "input_files": [],
+      "review_instruction": "Pick A or B based only on the rubric. Do not infer which output came from the skill.",
+      "rubric": [
+        {
+          "id": "has-ir-first",
+          "description": "Output requires Skill IR before packaging.",
+          "weight": 1.0
+        },
+        {
+          "id": "has-core-ir-fields",
+          "description": "Output lists core IR contract fields.",
+          "weight": 1.0
+        },
+        {
+          "id": "has-targets",
+          "description": "Output names requested runtime targets.",
+          "weight": 1.0
+        },
+        {
+          "id": "has-semantic-preservation",
+          "description": "Output says targets must preserve the capability contract.",
+          "weight": 1.0
+        }
+      ],
+      "variant_a": {
+        "blind_id": "ir-before-packaging:A",
+        "output": "Export folders for each platform and adjust files as needed."
+      },
+      "variant_b": {
+        "blind_id": "ir-before-packaging:B",
+        "output": "Export Skill IR first with job_to_be_done, trigger_surface, workflow, resources, eval_plan, risk, and governance. Then compile or package targets from the IR so OpenAI, Claude, Agent Skills, and generic packages preserve the same capability contract."
+      },
+      "decision_state": {
+        "status": "awaiting-decision",
+        "winner_variant_recorded": false,
+        "confidence_recorded": false,
+        "reason_recorded": false,
+        "blocking_reason": "Decision template exists but this row is still blank."
+      }
+    },
+    {
+      "case_id": "near-neighbor-boundary",
+      "prompt": "I only need a one-off summary of these notes, not a reusable process.",
+      "input_files": [],
+      "review_instruction": "Pick A or B based only on the rubric. Do not infer which output came from the skill.",
+      "rubric": [
+        {
+          "id": "declines-skill",
+          "description": "Output refuses unnecessary skill creation for one-off work.",
+          "weight": 1.0
+        },
+        {
+          "id": "names-near-neighbor",
+          "description": "Output labels the request as near-neighbor instead of owned work.",
+          "weight": 1.0
+        },
+        {
+          "id": "requires-reuse-signal",
+          "description": "Output asks for repeat-use evidence before packaging.",
+          "weight": 1.0
+        }
+      ],
+      "variant_a": {
+        "blind_id": "near-neighbor-boundary:A",
+        "output": "Create a SKILL.md and a reusable workflow anyway so future notes can use it."
+      },
+      "variant_b": {
+        "blind_id": "near-neighbor-boundary:B",
+        "output": "Do not create a skill for this one-off request. Treat it as a near-neighbor: answer the summary directly unless the user confirms repeated use, shared ownership, or a reusable output contract."
+      },
+      "decision_state": {
+        "status": "awaiting-decision",
+        "winner_variant_recorded": false,
+        "confidence_recorded": false,
+        "reason_recorded": false,
+        "blocking_reason": "Decision template exists but this row is still blank."
+      }
+    },
+    {
+      "case_id": "file-backed-governed-package",
+      "prompt": "Turn the attached release brief source into a governed skill package.",
+      "input_files": [
+        "fixtures/release-brief-source.md"
+      ],
+      "review_instruction": "Pick A or B based only on the rubric. Do not infer which output came from the skill.",
+      "rubric": [
+        {
+          "id": "uses-file-backed-evidence",
+          "description": "Output names file-backed source evidence.",
+          "weight": 1.0
+        },
+        {
+          "id": "has-governance",
+          "description": "Output preserves governed ownership metadata.",
+          "weight": 1.0
+        },
+        {
+          "id": "has-output-contract",
+          "description": "Output preserves output and rollback boundaries.",
+          "weight": 1.0
+        },
+        {
+          "id": "has-trust-and-scorecard",
+          "description": "Output requires trust and output scorecard artifacts.",
+          "weight": 1.0
+        },
+        {
+          "id": "does-not-invent-evidence",
+          "description": "Output forbids invented launch evidence.",
+          "weight": 1.0
+        }
+      ],
+      "variant_a": {
+        "blind_id": "file-backed-governed-package:A",
+        "output": "Draft a release announcement with the changelog, support notes, and owner name."
+      },
+      "variant_b": {
+        "blind_id": "file-backed-governed-package:B",
+        "output": "Use the file-backed fixture as source evidence, then create a governed skill package with SKILL.md, agents/interface.yaml, owner, review cadence, input_files, output contract, rollback boundary, trust report, and reports/output_quality_scorecard.md. Mark missing launch metrics as missing evidence instead of inventing them."
+      },
+      "decision_state": {
+        "status": "awaiting-decision",
+        "winner_variant_recorded": false,
+        "confidence_recorded": false,
+        "reason_recorded": false,
+        "blocking_reason": "Decision template exists but this row is still blank."
+      }
+    }
+  ],
+  "failures": []
+}
@@ -0,0 +1,138 @@
+# Output Review Kit
+
+This reviewer-facing packet contains the blind A/B cases, decision fields, and command flow. It intentionally does not expose the answer key.
+
+## Summary
+
+- cases: `5`
+- ready for adjudication: `0`
+- pending decisions: `5`
+- invalid decisions: `0`
+- reviewer metadata present: `false`
+- answer key hidden: `true`
+- answer key path exposed: `false`
+
+## Review Flow
+
+1. Open reports/output_blind_review_pack.md or this kit and compare Variant A vs Variant B for each case.
+2. Record choices in reports/output_review_decisions.json without opening the answer key.
+3. Use winner_variant A or B, confidence from 0 to 1, and a short reason for every case.
+4. Run python3 scripts/yao.py output-review after choices are recorded.
+5. Refresh python3 scripts/yao.py review-studio . before asking for release approval.
+
+## Required Fields
+
+- `reviewer`: Human reviewer name or review group.
+- `reviewed_at`: Review date or timestamp.
+- `winner_variant`: A or B for every case.
+- `confidence`: Optional numeric confidence from 0 to 1.
+- `reason`: Short rationale based on the rubric, not on hidden labels.
+
+## Privacy Contract
+
+- The answer key is intentionally withheld from this kit.
+- Do not inspect hidden labels or expected winners before decisions are recorded.
+- Do not paste private user data into decision reasons.
+- Pending decisions must stay pending instead of being counted as human agreement.
+
+## Decision States
+
+| Case | State | Winner | Confidence | Reason | Blocking Reason |
+| --- | --- | --- | --- | --- | --- |
+| `skill-package-contract` | `awaiting-decision` | `false` | `false` | `false` | Decision template exists but this row is still blank. |
+| `output-eval-expectation` | `awaiting-decision` | `false` | `false` | `false` | Decision template exists but this row is still blank. |
+| `ir-before-packaging` | `awaiting-decision` | `false` | `false` | `false` | Decision template exists but this row is still blank. |
+| `near-neighbor-boundary` | `awaiting-decision` | `false` | `false` | `false` | Decision template exists but this row is still blank. |
+| `file-backed-governed-package` | `awaiting-decision` | `false` | `false` | `false` | Decision template exists but this row is still blank. |
+
+## Blind Cases
+
+### skill-package-contract
+
+Prompt: Turn this repeated workflow into a reusable team skill package.
+
+Rubric:
+- `has-entrypoint` (1.0): Output names the routeable Skill entrypoint.
+- `has-interface` (1.0): Output includes neutral interface metadata.
+- `has-report` (1.0): Output points reviewer to generated overview report.
+- `has-resource-boundary` (1.0): Output preserves entrypoint/resource separation.
+
+#### Variant A
+
+Create a routeable SKILL.md, agents/interface.yaml, reports/skill-overview.html, and a focused eval plan. Keep the root entrypoint lean, move durable guidance into references/, scripts into scripts/, and evidence into reports/.
+
+#### Variant B
+
+I can write a prompt for that workflow and include a short checklist.
+
+### output-eval-expectation
+
+Prompt: Upgrade this production skill so we know whether the generated output is better than baseline.
+
+Rubric:
+- `has-baseline-comparison` (1.0): Output explicitly compares with-skill and baseline outputs.
+- `has-assertion-grading` (1.0): Output includes objective assertion grading.
+- `has-scorecard` (1.0): Output produces a scorecard report path.
+- `has-failure-taxonomy` (1.0): Output records failure taxonomy.
+
+#### Variant A
+
+Add Output Eval Lab cases with baseline_output and with_skill_output, run assertion grading, report with-skill vs baseline pass-rate delta, and record failure taxonomy plus next fixes in reports/output_quality_scorecard.md.
+
+#### Variant B
+
+Add more examples and run the trigger eval again.
+
+### ir-before-packaging
+
+Prompt: Package this skill for OpenAI, Claude, Agent Skills, and generic targets.
+
+Rubric:
+- `has-ir-first` (1.0): Output requires Skill IR before packaging.
+- `has-core-ir-fields` (1.0): Output lists core IR contract fields.
+- `has-targets` (1.0): Output names requested runtime targets.
+- `has-semantic-preservation` (1.0): Output says targets must preserve the capability contract.
+
+#### Variant A
+
+Export folders for each platform and adjust files as needed.
+
+#### Variant B
+
+Export Skill IR first with job_to_be_done, trigger_surface, workflow, resources, eval_plan, risk, and governance. Then compile or package targets from the IR so OpenAI, Claude, Agent Skills, and generic packages preserve the same capability contract.
+
+### near-neighbor-boundary
+
+Prompt: I only need a one-off summary of these notes, not a reusable process.
+
+Rubric:
+- `declines-skill` (1.0): Output refuses unnecessary skill creation for one-off work.
+- `names-near-neighbor` (1.0): Output labels the request as near-neighbor instead of owned work.
+- `requires-reuse-signal` (1.0): Output asks for repeat-use evidence before packaging.
+
+#### Variant A
+
+Create a SKILL.md and a reusable workflow anyway so future notes can use it.
+
+#### Variant B
+
+Do not create a skill for this one-off request. Treat it as a near-neighbor: answer the summary directly unless the user confirms repeated use, shared ownership, or a reusable output contract.
+
+### file-backed-governed-package
+
+Prompt: Turn the attached release brief source into a governed skill package.
+
+Rubric:
+- `uses-file-backed-evidence` (1.0): Output names file-backed source evidence.
+- `has-governance` (1.0): Output preserves governed ownership metadata.
+- `has-output-contract` (1.0): Output preserves output and rollback boundaries.
+- `has-trust-and-scorecard` (1.0): Output requires trust and output scorecard artifacts.
+- `does-not-invent-evidence` (1.0): Output forbids invented launch evidence.
+
+#### Variant A
+
+Draft a release announcement with the changelog, support notes, and owner name.
+
+#### Variant B
+
+Use the file-backed fixture as source evidence, then create a governed skill package with SKILL.md, agents/interface.yaml, owner, review cadence, input_files, output contract, rollback boundary, trust report, and reports/output_quality_scorecard.md. Mark missing launch metrics as missing evidence instead of inventing them.
@@ -5,7 +5,7 @@
  "root": ".",
  "summary": {
    "target_python": "3.11",
-    "file_count": 152,
+    "file_count": 154,
    "issue_count": 0,
    "syntax_error_count": 0,
    "fstring_311_violation_count": 0,
@@ -166,6 +166,12 @@
      "issue_count": 0,
      "issues": []
    },
+    {
+      "path": "scripts/prepare_output_review_kit.py",
+      "ok": true,
+      "issue_count": 0,
+      "issues": []
+    },
    {
      "path": "scripts/prepare_world_class_submission_kit.py",
      "ok": true,
@@ -706,6 +712,12 @@
      "issue_count": 0,
      "issues": []
    },
+    {
+      "path": "tests/verify_output_review_kit.py",
+      "ok": true,
+      "issue_count": 0,
+      "issues": []
+    },
    {
      "path": "tests/verify_output_risk_profile.py",
      "ok": true,
@@ -6,7 +6,7 @@ Generated at: `2026-06-13`

 - decision: `pass`
 - target python: `3.11`
- files scanned: `152`
+- files scanned: `154`
 - issues: `0`
 - syntax errors: `0`
 - f-string 3.11 violations: `0`
@@ -21,7 +21,7 @@
    "trust_level": "local",
    "license": "MIT",
    "checksums": {
-      "package_sha256": "68a0228b56de8c9e9e94d64d7c1e386bf3dff55c0f6228b65f91e75d02525926",
+      "package_sha256": "33ca97e8df3e3c75228595ab8bd06d65c697ef708723a0a53637bd8f22bbcfda",
      "archive_sha256": "cb1ac8c62bc5dee942a88f18eb5d71c93ae58f854ac19f3d912f9c9c3a4ec24e"
    },
    "compatibility": {
@@ -78,7 +78,7 @@
          "vscode"
        ],
        "package_metadata": "registry/packages/yao-meta-skill.json",
-        "package_sha256": "68a0228b56de8c9e9e94d64d7c1e386bf3dff55c0f6228b65f91e75d02525926"
+        "package_sha256": "33ca97e8df3e3c75228595ab8bd06d65c697ef708723a0a53637bd8f22bbcfda"
      }
    ]
  },
@@ -6,7 +6,7 @@
 - Maturity: `governed`
 - Owner: `Yao Team`
 - License: `MIT`
- Package SHA256: `68a0228b56de8c9e9e94d64d7c1e386bf3dff55c0f6228b65f91e75d02525926`
+- Package SHA256: `33ca97e8df3e3c75228595ab8bd06d65c697ef708723a0a53637bd8f22bbcfda`
 - Archive SHA256: `cb1ac8c62bc5dee942a88f18eb5d71c93ae58f854ac19f3d912f9c9c3a4ec24e`
 - Install simulated: `True`

@@ -2,8 +2,8 @@
  "ok": true,
  "skill_dir": ".",
  "summary": {
-    "scanned_files": 176,
-    "script_count": 92,
+    "scanned_files": 177,
+    "script_count": 93,
    "internal_module_count": 17,
    "secret_findings": 0,
    "dependency_files": [
@@ -12,18 +12,18 @@
    "network_script_count": 3,
    "network_policy_covered_count": 3,
    "network_policy_missing_count": 0,
-    "file_write_script_count": 62,
+    "file_write_script_count": 63,
    "permission_required_count": 3,
    "permission_approved_count": 3,
    "permission_missing_count": 0,
    "permission_invalid_count": 0,
    "permission_expired_count": 0,
-    "help_smoke_checked_count": 75,
+    "help_smoke_checked_count": 76,
    "help_smoke_failed_count": 0,
    "interactive_script_count": 0,
    "package_hash_scope": "source-contract-without-generated-reports",
-    "package_hash_file_count": 176,
-    "package_sha256": "68a0228b56de8c9e9e94d64d7c1e386bf3dff55c0f6228b65f91e75d02525926"
+    "package_hash_file_count": 177,
+    "package_sha256": "33ca97e8df3e3c75228595ab8bd06d65c697ef708723a0a53637bd8f22bbcfda"
  },
  "failures": [],
  "warnings": [],
@@ -332,6 +332,20 @@
      "network_urls": [],
      "network_hosts": []
    },
+    {
+      "path": "scripts/prepare_output_review_kit.py",
+      "interface": "cli",
+      "interface_declared": false,
+      "interface_reason": "Default CLI classification; add SCRIPT_INTERFACE for internal modules.",
+      "has_argparse": true,
+      "has_main_guard": true,
+      "uses_input": false,
+      "uses_network": false,
+      "uses_file_write": true,
+      "uses_subprocess": false,
+      "network_urls": [],
+      "network_hosts": []
+    },
    {
      "path": "scripts/prepare_world_class_submission_kit.py",
      "interface": "cli",
@@ -1365,9 +1379,9 @@
  "help_smoke": {
    "enabled": true,
    "timeout_seconds": 5.0,
-    "candidate_count": 75,
-    "checked_count": 75,
-    "passed_count": 75,
+    "candidate_count": 76,
+    "checked_count": 76,
+    "passed_count": 76,
    "failed_count": 0,
    "skipped_count": 17,
    "failed_scripts": [],
@@ -1582,6 +1596,16 @@
        "stdout_excerpt": "usage: optimize_description.py [-h] --description-file DESCRIPTION_FILE\n                               [--baseline-description-file BASELINE_DESCRIPTION_FILE]\n                               --dev-cases DEV_CASES\n                            ",
        "stderr_excerpt": ""
      },
+      {
+        "path": "scripts/prepare_output_review_kit.py",
+        "command": "python3 scripts/prepare_output_review_kit.py --help",
+        "returncode": 0,
+        "timed_out": false,
+        "passed": true,
+        "has_help_text": true,
+        "stdout_excerpt": "usage: prepare_output_review_kit.py [-h] [--blind-pack-json BLIND_PACK_JSON]\n                                    [--blind-pack-md BLIND_PACK_MD]\n                                    [--decisions DECISIONS]\n                                   ",
+        "stderr_excerpt": ""
+      },
      {
        "path": "scripts/prepare_world_class_submission_kit.py",
        "command": "python3 scripts/prepare_world_class_submission_kit.py --help",
@@ -2260,6 +2284,7 @@
          "scripts/github_benchmark_scan.py",
          "scripts/import_telemetry_events.py",
          "scripts/init_skill.py",
+          "scripts/prepare_output_review_kit.py",
          "scripts/prepare_world_class_submission_kit.py",
          "scripts/probe_runtime_permissions.py",
          "scripts/promotion_checker.py",
@@ -1,22 +1,22 @@
 # Security Trust Report

 - OK: `True`
- Scanned files: `176`
- Scripts: `92`
+- Scanned files: `177`
+- Scripts: `93`
 - Internal script modules: `17`
 - Secret findings: `0`
 - Network-capable scripts: `3`
 - Network policy covered scripts: `3`
 - Network policy missing scripts: `0`
- File-write scripts: `62`
+- File-write scripts: `63`
 - Permission approvals: `3 / 3`
 - Permission approval gaps: `0`
- CLI help smoke checked: `75`
+- CLI help smoke checked: `76`
 - CLI help smoke failures: `0`
 - Interactive scripts: `0`
 - Package hash scope: `source-contract-without-generated-reports`
- Package hash files: `176`
- Package SHA256: `68a0228b56de8c9e9e94d64d7c1e386bf3dff55c0f6228b65f91e75d02525926`
+- Package hash files: `177`
+- Package SHA256: `33ca97e8df3e3c75228595ab8bd06d65c697ef708723a0a53637bd8f22bbcfda`

 ## Failures

@@ -54,8 +54,8 @@

 - Enabled: `True`
 - Timeout seconds: `5.0`
- Checked scripts: `75`
- Passed scripts: `75`
+- Checked scripts: `76`
+- Passed scripts: `76`
 - Failed scripts: `none`

 ## Script Surface
@@ -83,6 +83,7 @@
 | scripts/lint_skill.py | cli | False | True | True | False | False | False | False | Default CLI classification; add SCRIPT_INTERFACE for internal modules. |
 | scripts/local_output_eval_runner.py | cli | False | True | True | False | False | False | False | Default CLI classification; add SCRIPT_INTERFACE for internal modules. |
 | scripts/optimize_description.py | cli | False | True | True | False | False | False | False | Default CLI classification; add SCRIPT_INTERFACE for internal modules. |
+| scripts/prepare_output_review_kit.py | cli | False | True | True | False | False | True | False | Default CLI classification; add SCRIPT_INTERFACE for internal modules. |
 | scripts/prepare_world_class_submission_kit.py | cli | True | True | True | False | False | True | False | Prepares editable world-class evidence intake packets without counting drafts as accepted evidence. |
 | scripts/probe_runtime_permissions.py | cli | False | True | True | False | False | True | False | Default CLI classification; add SCRIPT_INTERFACE for internal modules. |
 | scripts/promotion_checker.py | cli | False | True | True | False | False | True | False | Default CLI classification; add SCRIPT_INTERFACE for internal modules. |
@@ -621,7 +621,7 @@
        <div class="two-col">
          <article class="panel">
            <h3><span data-lang="zh-CN">作用定位</span><span data-lang="en">Role</span></h3>
-            <ul class="list"><li><span data-lang="zh-CN">从工作流、提示词、对话记录、文档或笔记中创建、重构、评估和打包 agent skill；适用于新建 Skill、沉淀重复流程、改进现有 Skill、补充 eval 或团队复用打包。</span><span data-lang="en">Create, refactor, evaluate, and package agent skills from workflows, prompts, transcripts, docs, or notes. Use when asked to create a skill, turn a repeated process into a reusable skill, improve an existing skill, add evals, or package a skill for team reuse.</span></li><li><span data-lang="zh-CN">把一次性经验沉淀为可复用、可评估、可迁移的 Skill 包体。</span><span data-lang="en">把一次性经验沉淀为可复用、可评估、可迁移的 Skill 包体。</span></li><li><span data-lang="zh-CN">交付结果：SKILL.md, agents/interface.yaml, reports/skill-ir.json, reports/compiled_targets.md, reports/output_quality_scorecard.md, reports/output_execution_runs.md, reports/output_blind_review_pack.md, reports/output_blind_answer_key.json, reports/output_review_adjudication.md, reports/benchmark_reproducibility.md, reports/conformance_matrix.md, reports/security_trust_report.md, reports/runtime_permission_probes.md, reports/skill_atlas.html, reports/registry_audit.md, reports/package_verification.md, reports/install_simulation.md, reports/upgrade_check.md, reports/adoption_drift_report.md, reports/review_waivers.md, reports/world_class_evidence_plan.md, reports/world_class_evidence_ledger.md, reports/review_annotations.md, reports/review-studio.html, reports/skill-overview.html</span><span data-lang="en">交付结果：SKILL.md, agents/interface.yaml, reports/skill-ir.json, reports/compiled_targets.md, reports/output_quality_scorecard.md, reports/output_execution_runs.md, reports/output_blind_review_pack.md, reports/output_blind_answer_key.json, reports/output_review_adjudication.md, reports/benchmark_reproducibility.md, reports/conformance_matrix.md, reports/security_trust_report.md, reports/runtime_permission_probes.md, reports/skill_atlas.html, reports/registry_audit.md, reports/package_verification.md, reports/install_simulation.md, reports/upgrade_check.md, reports/adoption_drift_report.md, reports/review_waivers.md, reports/world_class_evidence_plan.md, reports/world_class_evidence_ledger.md, reports/review_annotations.md, reports/review-studio.html, reports/skill-overview.html</span></li></ul>
+            <ul class="list"><li><span data-lang="zh-CN">从工作流、提示词、对话记录、文档或笔记中创建、重构、评估和打包 agent skill；适用于新建 Skill、沉淀重复流程、改进现有 Skill、补充 eval 或团队复用打包。</span><span data-lang="en">Create, refactor, evaluate, and package agent skills from workflows, prompts, transcripts, docs, or notes. Use when asked to create a skill, turn a repeated process into a reusable skill, improve an existing skill, add evals, or package a skill for team reuse.</span></li><li><span data-lang="zh-CN">把一次性经验沉淀为可复用、可评估、可迁移的 Skill 包体。</span><span data-lang="en">把一次性经验沉淀为可复用、可评估、可迁移的 Skill 包体。</span></li><li><span data-lang="zh-CN">交付结果：SKILL.md, agents/interface.yaml, reports/skill-ir.json, reports/compiled_targets.md, reports/output_quality_scorecard.md, reports/output_execution_runs.md, reports/output_blind_review_pack.md, reports/output_review_kit.md, reports/output_review_adjudication.md, reports/benchmark_reproducibility.md, reports/output_blind_answer_key.json, reports/conformance_matrix.md, reports/security_trust_report.md, reports/runtime_permission_probes.md, reports/skill_atlas.html, reports/registry_audit.md, reports/package_verification.md, reports/install_simulation.md, reports/upgrade_check.md, reports/adoption_drift_report.md, reports/review_waivers.md, reports/world_class_evidence_plan.md, reports/world_class_evidence_ledger.md, reports/review_annotations.md, reports/review-studio.html, reports/skill-overview.html</span><span data-lang="en">交付结果：SKILL.md, agents/interface.yaml, reports/skill-ir.json, reports/compiled_targets.md, reports/output_quality_scorecard.md, reports/output_execution_runs.md, reports/output_blind_review_pack.md, reports/output_review_kit.md, reports/output_review_adjudication.md, reports/benchmark_reproducibility.md, reports/output_blind_answer_key.json, reports/conformance_matrix.md, reports/security_trust_report.md, reports/runtime_permission_probes.md, reports/skill_atlas.html, reports/registry_audit.md, reports/package_verification.md, reports/install_simulation.md, reports/upgrade_check.md, reports/adoption_drift_report.md, reports/review_waivers.md, reports/world_class_evidence_plan.md, reports/world_class_evidence_ledger.md, reports/review_annotations.md, reports/review-studio.html, reports/skill-overview.html</span></li></ul>
          </article>
          <figure class="chart-figure" data-chart="flow"><svg viewBox="0 0 620 170" role="img" aria-label="交付流程"><text x="20" y="28" class="chart-title">交付流程</text><path d="M188 93 H248 M398 93 H458" class="chart-line"/><g><rect x="38" y="56" width="150" height="74" rx="8" fill="#F6F8FB" stroke="#e8e6dc"/><text x="113" y="99" text-anchor="middle">输入材料</text></g><g><rect x="248" y="56" width="150" height="74" rx="8" fill="#F6F8FB" stroke="#e8e6dc"/><text x="323" y="99" text-anchor="middle">Skill 包体</text></g><g><rect x="458" y="56" width="150" height="74" rx="8" fill="#F6F8FB" stroke="#e8e6dc"/><text x="533" y="99" text-anchor="middle">可复用能力</text></g></svg><figcaption>交付流程把用户输入、生成的包体和可复用能力放在一条线上。</figcaption></figure>
        </div>
@@ -771,7 +771,7 @@
          <p><span data-lang="zh-CN">让 reviewer 快速确认关键文件、目录和资产分布。</span><span data-lang="en">Lets reviewers confirm key files, directories, and asset distribution quickly.</span></p>
        </div>
        <div class="two-col">
-          <figure class="chart-figure" data-chart="asset_donut"><svg viewBox="0 0 430 270" role="img" aria-label="资产分布"><text x="20" y="30" class="chart-title">资产分布</text><circle cx="130" cy="130" r="70" fill="none" stroke="#1B365D" stroke-width="24" stroke-dasharray="0.8 99.2" stroke-dashoffset="0.0" pathLength="100" transform="rotate(-90 130 130)"/><circle cx="130" cy="130" r="70" fill="none" stroke="#2D5A8A" stroke-width="24" stroke-dasharray="0.8 99.2" stroke-dashoffset="-0.8" pathLength="100" transform="rotate(-90 130 130)"/><circle cx="130" cy="130" r="70" fill="none" stroke="#D0DCE9" stroke-width="24" stroke-dasharray="0.8 99.2" stroke-dashoffset="-1.6" pathLength="100" transform="rotate(-90 130 130)"/><circle cx="130" cy="130" r="70" fill="none" stroke="#E4ECF5" stroke-width="24" stroke-dasharray="0.8 99.2" stroke-dashoffset="-2.3" pathLength="100" transform="rotate(-90 130 130)"/><circle cx="130" cy="130" r="70" fill="none" stroke="#e8e6dc" stroke-width="24" stroke-dasharray="25.0 75.0" stroke-dashoffset="-3.1" pathLength="100" transform="rotate(-90 130 130)"/><circle cx="130" cy="130" r="70" fill="none" stroke="#504e49" stroke-width="24" stroke-dasharray="71.9 28.1" stroke-dashoffset="-28.1" pathLength="100" transform="rotate(-90 130 130)"/><text x="130" y="136" text-anchor="middle">353项</text><text x="235" y="78">SKILL.md</text><text x="235" y="100">README.md</text><text x="235" y="122">agents/interface.yaml</text><text x="235" y="144">manifest.json</text><text x="235" y="166">references</text><text x="235" y="188">scripts</text></svg><figcaption>资产分布图展示当前包体的文件和目录重心。</figcaption></figure>
+          <figure class="chart-figure" data-chart="asset_donut"><svg viewBox="0 0 430 270" role="img" aria-label="资产分布"><text x="20" y="30" class="chart-title">资产分布</text><circle cx="130" cy="130" r="70" fill="none" stroke="#1B365D" stroke-width="24" stroke-dasharray="0.8 99.2" stroke-dashoffset="0.0" pathLength="100" transform="rotate(-90 130 130)"/><circle cx="130" cy="130" r="70" fill="none" stroke="#2D5A8A" stroke-width="24" stroke-dasharray="0.8 99.2" stroke-dashoffset="-0.8" pathLength="100" transform="rotate(-90 130 130)"/><circle cx="130" cy="130" r="70" fill="none" stroke="#D0DCE9" stroke-width="24" stroke-dasharray="0.8 99.2" stroke-dashoffset="-1.6" pathLength="100" transform="rotate(-90 130 130)"/><circle cx="130" cy="130" r="70" fill="none" stroke="#E4ECF5" stroke-width="24" stroke-dasharray="0.8 99.2" stroke-dashoffset="-2.3" pathLength="100" transform="rotate(-90 130 130)"/><circle cx="130" cy="130" r="70" fill="none" stroke="#e8e6dc" stroke-width="24" stroke-dasharray="24.8 75.2" stroke-dashoffset="-3.1" pathLength="100" transform="rotate(-90 130 130)"/><circle cx="130" cy="130" r="70" fill="none" stroke="#504e49" stroke-width="24" stroke-dasharray="72.1 27.9" stroke-dashoffset="-27.9" pathLength="100" transform="rotate(-90 130 130)"/><text x="130" y="136" text-anchor="middle">356项</text><text x="235" y="78">SKILL.md</text><text x="235" y="100">README.md</text><text x="235" y="122">agents/interface.yaml</text><text x="235" y="144">manifest.json</text><text x="235" y="166">references</text><text x="235" y="188">scripts</text></svg><figcaption>资产分布图展示当前包体的文件和目录重心。</figcaption></figure>
          <table>
            <thead><tr><th><span data-lang="zh-CN">路径</span><span data-lang="en">Path</span></th><th><span data-lang="zh-CN">作用</span><span data-lang="en">Role</span></th><th><span data-lang="zh-CN">类型</span><span data-lang="en">Type</span></th></tr></thead>
            <tbody><tr><td>SKILL.md</td><td><span data-lang="zh-CN">Skill 入口文件</span><span data-lang="en">Skill entrypoint</span></td><td><span data-lang="zh-CN">文件</span><span data-lang="en">file</span></td></tr><tr><td>README.md</td><td><span data-lang="zh-CN">人类可读使用说明</span><span data-lang="en">Human-readable usage guide</span></td><td><span data-lang="zh-CN">文件</span><span data-lang="en">file</span></td></tr><tr><td>agents/interface.yaml</td><td><span data-lang="zh-CN">跨平台接口元数据</span><span data-lang="en">Neutral interface metadata</span></td><td><span data-lang="zh-CN">文件</span><span data-lang="en">file</span></td></tr><tr><td>manifest.json</td><td><span data-lang="zh-CN">生命周期与打包元数据</span><span data-lang="en">Lifecycle and portability metadata</span></td><td><span data-lang="zh-CN">文件</span><span data-lang="en">file</span></td></tr><tr><td>references</td><td><span data-lang="zh-CN">扩展指导与复用资料</span><span data-lang="en">Extended guidance and reusable notes</span></td><td><span data-lang="zh-CN">目录</span><span data-lang="en">folder</span></td></tr><tr><td>scripts</td><td><span data-lang="zh-CN">确定性脚本或本地工具</span><span data-lang="en">Deterministic helpers or local tooling</span></td><td><span data-lang="zh-CN">目录</span><span data-lang="en">folder</span></td></tr><tr><td>evals</td><td><span data-lang="zh-CN">触发与质量检查</span><span data-lang="en">Trigger and quality checks</span></td><td><span data-lang="zh-CN">目录</span><span data-lang="en">folder</span></td></tr><tr><td>reports</td><td><span data-lang="zh-CN">生成的证据与总结报告</span><span data-lang="en">Generated evidence and overview artifacts</span></td><td><span data-lang="zh-CN">目录</span><span data-lang="en">folder</span></td></tr></tbody>
@@ -16,9 +16,10 @@
      "reports/output_quality_scorecard.md",
      "reports/output_execution_runs.md",
      "reports/output_blind_review_pack.md",
-      "reports/output_blind_answer_key.json",
+      "reports/output_review_kit.md",
      "reports/output_review_adjudication.md",
      "reports/benchmark_reproducibility.md",
+      "reports/output_blind_answer_key.json",
      "reports/conformance_matrix.md",
      "reports/security_trust_report.md",
      "reports/runtime_permission_probes.md",
@@ -436,7 +437,7 @@
        "path": "scripts",
        "label": "Deterministic helpers or local tooling",
        "kind": "folder",
-        "file_count": 92
+        "file_count": 93
      },
      {
        "path": "evals",
@@ -448,10 +449,10 @@
        "path": "reports",
        "label": "Generated evidence and overview artifacts",
        "kind": "folder",
-        "file_count": 197
+        "file_count": 199
      }
    ],
-    "file_count": 353,
+    "file_count": 356,
    "folder_count": 4,
    "distribution": [
      {
@@ -476,7 +477,7 @@
      },
      {
        "label": "scripts",
-        "value": 92
+        "value": 93
      },
      {
        "label": "evals",
@@ -484,7 +485,7 @@
      },
      {
        "label": "reports",
-        "value": 197
+        "value": 199
      }
    ]
  },
@@ -605,7 +606,7 @@
      "path": "scripts",
      "label": "Deterministic helpers or local tooling",
      "kind": "folder",
-      "file_count": 92
+      "file_count": 93
    },
    {
      "path": "evals",
@@ -617,7 +618,7 @@
      "path": "reports",
      "label": "Generated evidence and overview artifacts",
      "kind": "folder",
-      "file_count": 197
+      "file_count": 199
    }
  ],
  "strengths": [
@@ -865,6 +866,30 @@
    "pair_count": 5,
    "answer_key_separate": true
  },
+  "output_review_kit": {
+    "ok": true,
+    "summary": {
+      "case_count": 5,
+      "ready_for_adjudication_count": 0,
+      "pending_decision_count": 5,
+      "invalid_decision_count": 0,
+      "reviewer_metadata_present": false,
+      "answer_key_hidden": true,
+      "answer_key_path_exposed": false,
+      "ready_to_run_adjudication": false,
+      "failure_count": 0
+    },
+    "artifacts": {
+      "reviewer_kit_json": "reports/output_review_kit.json",
+      "reviewer_kit_markdown": "reports/output_review_kit.md",
+      "blind_pack_json": "reports/output_blind_review_pack.json",
+      "blind_pack_markdown": "reports/output_blind_review_pack.md",
+      "decisions": "reports/output_review_decisions.json",
+      "adjudication_json": "reports/output_review_adjudication.json",
+      "adjudication_markdown": "reports/output_review_adjudication.md"
+    },
+    "failures": []
+  },
  "output_review_adjudication": {
    "ok": true,
    "summary": {
@@ -909,9 +934,9 @@
      "world_class_task_count": 4,
      "world_class_ledger_pending_count": 4,
      "working_tree_dirty": true,
-      "changed_file_count": 56
+      "changed_file_count": 59
    },
-    "commit": "30539bed7a29dd88c2f289dd0c115da9834dd5d9",
+    "commit": "ecdb16dd90025bcca3051be4c6a1b5b723c7ec75",
    "missing_artifacts": [],
    "limitations": [
      "Local command-runner evidence is reproducible but does not replace provider-backed model holdout evidence.",
@@ -986,8 +1011,8 @@
    "failures": []
  },
  "trust_security": {
-    "scanned_files": 176,
-    "script_count": 92,
+    "scanned_files": 177,
+    "script_count": 93,
    "internal_module_count": 17,
    "secret_findings": 0,
    "dependency_files": [
@@ -996,18 +1021,18 @@
    "network_script_count": 3,
    "network_policy_covered_count": 3,
    "network_policy_missing_count": 0,
-    "file_write_script_count": 62,
+    "file_write_script_count": 63,
    "permission_required_count": 3,
    "permission_approved_count": 3,
    "permission_missing_count": 0,
    "permission_invalid_count": 0,
    "permission_expired_count": 0,
-    "help_smoke_checked_count": 75,
+    "help_smoke_checked_count": 76,
    "help_smoke_failed_count": 0,
    "interactive_script_count": 0,
    "package_hash_scope": "source-contract-without-generated-reports",
-    "package_hash_file_count": 176,
-    "package_sha256": "6298aa08d70a1e36956df32bcbb7a0bc778e8125f5449630cebf966b1c06a5f7"
+    "package_hash_file_count": 177,
+    "package_sha256": "7645bfc56829c3d6888e792b170b2e734d041c930c31f7f285c0d59e49d7de1b"
  },
  "skill_atlas": {
    "skill_count": 12,
@@ -1045,7 +1070,7 @@
      "trust_level": "local",
      "license": "MIT",
      "checksums": {
-        "package_sha256": "6298aa08d70a1e36956df32bcbb7a0bc778e8125f5449630cebf966b1c06a5f7",
+        "package_sha256": "7645bfc56829c3d6888e792b170b2e734d041c930c31f7f285c0d59e49d7de1b",
        "archive_sha256": "cb1ac8c62bc5dee942a88f18eb5d71c93ae58f854ac19f3d912f9c9c3a4ec24e"
      },
      "compatibility": {
@@ -1082,7 +1107,7 @@
        "install_simulated": true,
        "install_simulation": "reports/install_simulation.json"
      },
-      "generated_at": "2026-06-14"
+      "generated_at": "2026-06-13"
    },
    "failures": [],
    "warnings": []
@@ -1105,7 +1130,7 @@
    "ok": true,
    "summary": {
      "archive_present": true,
-      "archive_entry_count": 563,
+      "archive_entry_count": 567,
      "archive_extracted": true,
      "entrypoint_loaded": true,
      "manifest_loaded": true,
@@ -1177,7 +1202,7 @@
        {
          "field": "package_sha256",
          "from": "0000000000000000000000000000000000000000000000000000000000000000",
-          "to": "6298aa08d70a1e36956df32bcbb7a0bc778e8125f5449630cebf966b1c06a5f7"
+          "to": "7645bfc56829c3d6888e792b170b2e734d041c930c31f7f285c0d59e49d7de1b"
        }
      ]
    },
@@ -1349,8 +1374,9 @@
        "current": "0/5 decisions; pending 5",
        "objective": "Record real blind A/B reviewer decisions before claiming human output review completion.",
        "runbook": [
+          "python3 scripts/yao.py output-review-kit --write-template",
+          "Open reports/output_review_kit.md and choose A or B for each pair without opening the answer key.",
          "python3 scripts/adjudicate_output_review.py --write-template",
-          "Open reports/output_blind_review_pack.md and choose A or B for each pair without opening the answer key.",
          "Edit reports/output_review_decisions.json with winner_variant values and reviewer metadata.",
          "python3 scripts/yao.py output-review",
          "python3 scripts/yao.py skill-os2-audit . --generated-at <YYYY-MM-DD>",
@@ -1365,6 +1391,7 @@
        ],
        "evidence_artifacts": [
          "reports/output_blind_review_pack.md",
+          "reports/output_review_kit.md",
          "reports/output_review_decisions.json",
          "reports/output_review_adjudication.json",
          "reports/output_review_adjudication.md",
@@ -1556,6 +1583,7 @@
        ],
        "evidence_artifacts": [
          "reports/output_blind_review_pack.md",
+          "reports/output_review_kit.md",
          "reports/output_review_decisions.json",
          "reports/output_review_adjudication.json",
          "reports/output_review_adjudication.md",
@@ -85,6 +85,7 @@
          "scripts/lint_skill.py",
          "scripts/local_output_eval_runner.py",
          "scripts/optimize_description.py",
+          "scripts/prepare_output_review_kit.py",
          "scripts/prepare_world_class_submission_kit.py",
          "scripts/probe_runtime_permissions.py",
          "scripts/promotion_checker.py",
@@ -197,7 +197,7 @@
      "key": "trust-security",
      "label": "Trust Security",
      "status": "pass",
-      "current": "secrets 0; scripts 92; help failures 0",
+      "current": "secrets 0; scripts 93; help failures 0",
      "target": "Secrets, scripts, dependencies, permissions, and package hash are reviewable",
      "evidence": [
        {
@@ -23,7 +23,7 @@ Generated at: `2026-06-13`
 | Human Adjudication | human-required | 0/5 decisions; pending 5 | Real reviewer decisions recorded before claiming output review completion | Record real A/B choices in the decision template, then regenerate adjudication. |
 | Benchmark Reproducibility | pass | artifacts 20; missing 0; failures 3 | Public methodology, reproducible commands, required artifacts, and failure disclosure are machine-checkable | Keep the manifest current with every benchmark, package, and release evidence change. |
 | Runtime Conformance | pass | 5/5 targets pass | Target package structure, metadata, relative paths, and degradation notes pass | Keep target conformance fixtures updated as platform contracts change. |
-| Trust Security | pass | secrets 0; scripts 92; help failures 0 | Secrets, scripts, dependencies, permissions, and package hash are reviewable | Keep high-permission approvals scoped, expiring, and target-mapped. |
+| Trust Security | pass | secrets 0; scripts 93; help failures 0 | Secrets, scripts, dependencies, permissions, and package hash are reviewable | Keep high-permission approvals scoped, expiring, and target-mapped. |
 | Permission Metadata | pass | 4/4 target probes pass; metadata fallback 4 | Packaged adapters expose explicit permission metadata and residual risks | Preserve residual-risk notes until real native enforcement exists. |
 | Native Permission Enforcement | external-required | native-enforced targets 0 | At least one target/client enforces approved permissions at runtime | Integrate a real client or installer runtime guard before claiming native permission enforcement. |
 | Skill Atlas | pass | 12 skills; actionable collisions 0 | Workspace catalog, route overlap, stale/owner gaps, drift, and no-route opportunities | Feed real drift data into Atlas once client telemetry is installed. |
@@ -122,7 +122,7 @@
      "label": "Trust Security",
      "status": "pass",
      "objective": "Scripts, dependencies, permissions, secrets, and package hash are reviewable for team distribution.",
-      "current": "92 scripts; secrets 0; help failures 0",
+      "current": "93 scripts; secrets 0; help failures 0",
      "command": "python3 scripts/yao.py trust .",
      "test": "python3 tests/verify_trust_check.py",
      "evidence": [
@@ -21,7 +21,7 @@ This report maps the Skill OS 2.0 upgrade blueprint to concrete local artifacts,
 | Skill IR | `pass` | schema 2.0.0; targets 5 | `python3 scripts/yao.py skill-ir .` | `python3 tests/verify_skill_ir.py` |
 | Output Eval Lab | `pass` | 5 cases; delta 100.0; execution 10 | `python3 scripts/yao.py output-exec . && python3 scripts/yao.py output-review .` | `python3 tests/verify_output_eval_lab.py` |
 | Runtime Conformance | `pass` | 5/5 targets pass | `python3 scripts/yao.py conformance .` | `python3 tests/verify_conformance_suite.py` |
-| Trust Security | `pass` | 92 scripts; secrets 0; help failures 0 | `python3 scripts/yao.py trust .` | `python3 tests/verify_trust_check.py` |
+| Trust Security | `pass` | 93 scripts; secrets 0; help failures 0 | `python3 scripts/yao.py trust .` | `python3 tests/verify_trust_check.py` |
 | Skill Atlas | `pass` | 12 scanned skills; actionable collisions 0 | `python3 scripts/yao.py skill-atlas --workspace-root .` | `python3 tests/verify_skill_atlas.py` |
 | Registry Distribution | `pass` | archive entries 563; install failures 0 | `python3 scripts/yao.py package . --platform openai --platform claude --platform generic --platform vscode --output-dir dist --zip && python3 scripts/yao.py registry-audit .` | `python3 tests/verify_registry_audit.py` |
 | Review Studio | `pass` | 16 gates; decision review; warnings 3 | `python3 scripts/yao.py review-studio .` | `python3 tests/verify_review_studio.py` |
@@ -75,7 +75,7 @@
      {
        "field": "package_sha256",
        "from": "0000000000000000000000000000000000000000000000000000000000000000",
-        "to": "68a0228b56de8c9e9e94d64d7c1e386bf3dff55c0f6228b65f91e75d02525926"
+        "to": "33ca97e8df3e3c75228595ab8bd06d65c697ef708723a0a53637bd8f22bbcfda"
      }
    ]
  },
@@ -6,7 +6,7 @@
  "summary": {
    "ledger_ready_to_claim_world_class": false,
    "ledger_pending_count": 4,
-    "claim_surface_count": 69,
+    "claim_surface_count": 70,
    "violation_count": 0,
    "overclaim_guard_active": true,
    "decision": "claim-guard-pass-evidence-pending"
@@ -195,6 +195,10 @@
      "path": "reports/output_review_adjudication.md",
      "violation_count": 0
    },
+    {
+      "path": "reports/output_review_kit.md",
+      "violation_count": 0
+    },
    {
      "path": "reports/package_verification.md",
      "violation_count": 0
@@ -7,7 +7,7 @@ Generated at: `2026-06-13`
 - decision: `claim-guard-pass-evidence-pending`
 - ledger ready to claim world-class: `false`
 - ledger pending evidence: `4`
- claim surfaces scanned: `69`
+- claim surfaces scanned: `70`
 - violations: `0`
 - overclaim guard active: `true`

@@ -134,6 +134,7 @@
        ],
        "evidence_artifacts": [
          "reports/output_blind_review_pack.md",
+          "reports/output_review_kit.md",
          "reports/output_review_decisions.json",
          "reports/output_review_adjudication.json",
          "reports/output_review_adjudication.md",
@@ -107,6 +107,7 @@ This report validates the intake contract for human and external evidence. A val
  - reports/skill_os2_audit.json item human-adjudication status becomes pass
 - evidence_artifacts:
  - reports/output_blind_review_pack.md
+  - reports/output_review_kit.md
  - reports/output_review_decisions.json
  - reports/output_review_adjudication.json
  - reports/output_review_adjudication.md
@@ -95,6 +95,7 @@
      ],
      "evidence_artifacts": [
        "reports/output_blind_review_pack.md",
+        "reports/output_review_kit.md",
        "reports/output_review_decisions.json",
        "reports/output_review_adjudication.json",
        "reports/output_review_adjudication.md",
@@ -58,8 +58,9 @@
      "current": "0/5 decisions; pending 5",
      "objective": "Record real blind A/B reviewer decisions before claiming human output review completion.",
      "runbook": [
+        "python3 scripts/yao.py output-review-kit --write-template",
+        "Open reports/output_review_kit.md and choose A or B for each pair without opening the answer key.",
        "python3 scripts/adjudicate_output_review.py --write-template",
-        "Open reports/output_blind_review_pack.md and choose A or B for each pair without opening the answer key.",
        "Edit reports/output_review_decisions.json with winner_variant values and reviewer metadata.",
        "python3 scripts/yao.py output-review",
        "python3 scripts/yao.py skill-os2-audit . --generated-at <YYYY-MM-DD>",
@@ -74,6 +75,7 @@
      ],
      "evidence_artifacts": [
        "reports/output_blind_review_pack.md",
+        "reports/output_review_kit.md",
        "reports/output_review_decisions.json",
        "reports/output_review_adjudication.json",
        "reports/output_review_adjudication.md",
@@ -63,8 +63,9 @@ This report is an execution plan for the remaining world-class evidence gaps. It

 ### Runbook

+- `python3 scripts/yao.py output-review-kit --write-template`
+- Open reports/output_review_kit.md and choose A or B for each pair without opening the answer key.
 - `python3 scripts/adjudicate_output_review.py --write-template`
- Open reports/output_blind_review_pack.md and choose A or B for each pair without opening the answer key.
 - Edit reports/output_review_decisions.json with winner_variant values and reviewer metadata.
 - `python3 scripts/yao.py output-review`
 - `python3 scripts/yao.py skill-os2-audit . --generated-at <YYYY-MM-DD>`
@@ -81,6 +82,7 @@ This report is an execution plan for the remaining world-class evidence gaps. It
 ### Evidence Artifacts

 - `reports/output_blind_review_pack.md`
+- `reports/output_review_kit.md`
 - `reports/output_review_decisions.json`
 - `reports/output_review_adjudication.json`
 - `reports/output_review_adjudication.md`
@@ -245,6 +245,7 @@ def build_reviewer_checklist(
                "blind_pack_path": display_path(blind_pack_path),
                "decisions_path": display_path(decisions_path),
                "commands": {
+                    "prepare_review_kit": "python3 scripts/yao.py output-review-kit",
                    "write_template": "python3 scripts/adjudicate_output_review.py --write-template",
                    "adjudicate": "python3 scripts/yao.py output-review",
                    "refresh_review_studio": "python3 scripts/yao.py review-studio .",
@@ -40,6 +40,7 @@ DEFAULT_TARGETS = [
    "compiler-check",
    "output-eval-check",
    "output-execution-check",
+    "output-review-kit-check",
    "output-review-adjudication-check",
    "runtime-conformance-check",
    "runtime-permission-check",
@@ -0,0 +1,316 @@
+#!/usr/bin/env python3
+"""Prepare a reviewer-facing blind A/B output review kit."""
+
+import argparse
+import json
+from pathlib import Path
+from typing import Any
+
+from adjudicate_output_review import (
+    build_decision_template,
+    confidence_value,
+    decision_index,
+    display_path,
+    load_json,
+    normalize_variant,
+)
+
+
+ROOT = Path(__file__).resolve().parent.parent
+DEFAULT_BLIND_PACK_JSON = ROOT / "reports" / "output_blind_review_pack.json"
+DEFAULT_BLIND_PACK_MD = ROOT / "reports" / "output_blind_review_pack.md"
+DEFAULT_DECISIONS = ROOT / "reports" / "output_review_decisions.json"
+DEFAULT_OUTPUT_JSON = ROOT / "reports" / "output_review_kit.json"
+DEFAULT_OUTPUT_MD = ROOT / "reports" / "output_review_kit.md"
+
+
+def load_optional_decisions(path: Path) -> tuple[dict[str, Any], list[str]]:
+    if not path.exists():
+        return {"schema_version": "1.0", "decisions": []}, []
+    return load_json(path)
+
+
+def pairs_from_pack(blind_pack: dict[str, Any]) -> list[dict[str, Any]]:
+    pairs = blind_pack.get("pairs", [])
+    if not isinstance(pairs, list):
+        return []
+    return [item for item in pairs if isinstance(item, dict) and item.get("case_id")]
+
+
+def decision_state(decision: dict[str, Any] | None) -> dict[str, Any]:
+    if not decision:
+        return {
+            "status": "awaiting-decision",
+            "winner_variant_recorded": False,
+            "confidence_recorded": False,
+            "reason_recorded": False,
+            "blocking_reason": "No reviewer choice has been recorded yet.",
+        }
+    winner = normalize_variant(decision.get("winner_variant", decision.get("reviewer_winner_variant", decision.get("winner", ""))))
+    confidence, confidence_failure = confidence_value(decision.get("confidence"))
+    reason = str(decision.get("reason", "")).strip()
+    if not winner and confidence is None and not reason:
+        return {
+            "status": "awaiting-decision",
+            "winner_variant_recorded": False,
+            "confidence_recorded": False,
+            "reason_recorded": False,
+            "blocking_reason": "Decision template exists but this row is still blank.",
+        }
+    failures = []
+    if winner not in {"A", "B"}:
+        failures.append("winner_variant must be A or B")
+    if confidence_failure:
+        failures.append(confidence_failure)
+    if not reason:
+        failures.append("reason should explain the reviewer rationale")
+    if failures:
+        return {
+            "status": "needs-fix",
+            "winner_variant_recorded": winner in {"A", "B"},
+            "confidence_recorded": confidence is not None,
+            "reason_recorded": bool(reason),
+            "blocking_reason": "; ".join(failures),
+        }
+    return {
+        "status": "ready-for-adjudication",
+        "winner_variant_recorded": True,
+        "confidence_recorded": confidence is not None,
+        "reason_recorded": True,
+        "blocking_reason": "Reviewer choice is complete enough for adjudication.",
+    }
+
+
+def build_cases(blind_pack: dict[str, Any], decisions: dict[str, dict[str, Any]]) -> list[dict[str, Any]]:
+    cases = []
+    for pair in pairs_from_pack(blind_pack):
+        case_id = str(pair.get("case_id", ""))
+        rubric = []
+        for item in pair.get("rubric", []) if isinstance(pair.get("rubric", []), list) else []:
+            if isinstance(item, dict):
+                rubric.append(
+                    {
+                        "id": str(item.get("id", "")),
+                        "description": str(item.get("description", "")),
+                        "weight": float(item.get("weight", 1) or 0),
+                    }
+                )
+        cases.append(
+            {
+                "case_id": case_id,
+                "prompt": str(pair.get("prompt", "")),
+                "input_files": pair.get("input_files", []) if isinstance(pair.get("input_files", []), list) else [],
+                "review_instruction": str(pair.get("review_instruction", "")),
+                "rubric": rubric,
+                "variant_a": {
+                    "blind_id": str(pair.get("variant_a", {}).get("blind_id", "")) if isinstance(pair.get("variant_a"), dict) else "",
+                    "output": str(pair.get("variant_a", {}).get("output", "")) if isinstance(pair.get("variant_a"), dict) else "",
+                },
+                "variant_b": {
+                    "blind_id": str(pair.get("variant_b", {}).get("blind_id", "")) if isinstance(pair.get("variant_b"), dict) else "",
+                    "output": str(pair.get("variant_b", {}).get("output", "")) if isinstance(pair.get("variant_b"), dict) else "",
+                },
+                "decision_state": decision_state(decisions.get(case_id)),
+            }
+        )
+    return cases
+
+
+def build_summary(cases: list[dict[str, Any]], failures: list[str], reviewer: str, reviewed_at: str) -> dict[str, Any]:
+    status_counts: dict[str, int] = {}
+    for case in cases:
+        status = str(case.get("decision_state", {}).get("status", "awaiting-decision"))
+        status_counts[status] = status_counts.get(status, 0) + 1
+    ready_count = status_counts.get("ready-for-adjudication", 0)
+    pending_count = status_counts.get("awaiting-decision", 0)
+    invalid_count = status_counts.get("needs-fix", 0)
+    return {
+        "case_count": len(cases),
+        "ready_for_adjudication_count": ready_count,
+        "pending_decision_count": pending_count,
+        "invalid_decision_count": invalid_count,
+        "reviewer_metadata_present": bool(str(reviewer).strip() and str(reviewed_at).strip()),
+        "answer_key_hidden": True,
+        "answer_key_path_exposed": False,
+        "ready_to_run_adjudication": bool(cases) and ready_count == len(cases) and invalid_count == 0,
+        "failure_count": len(failures),
+    }
+
+
+def build_contract(blind_pack_md: Path, decisions_path: Path) -> dict[str, Any]:
+    return {
+        "reviewer_steps": [
+            f"Open {display_path(blind_pack_md)} or this kit and compare Variant A vs Variant B for each case.",
+            f"Record choices in {display_path(decisions_path)} without opening the answer key.",
+            "Use winner_variant A or B, confidence from 0 to 1, and a short reason for every case.",
+            "Run python3 scripts/yao.py output-review after choices are recorded.",
+            "Refresh python3 scripts/yao.py review-studio . before asking for release approval.",
+        ],
+        "required_fields": {
+            "reviewer": "Human reviewer name or review group.",
+            "reviewed_at": "Review date or timestamp.",
+            "winner_variant": "A or B for every case.",
+            "confidence": "Optional numeric confidence from 0 to 1.",
+            "reason": "Short rationale based on the rubric, not on hidden labels.",
+        },
+        "privacy_contract": [
+            "The answer key is intentionally withheld from this kit.",
+            "Do not inspect hidden labels or expected winners before decisions are recorded.",
+            "Do not paste private user data into decision reasons.",
+            "Pending decisions must stay pending instead of being counted as human agreement.",
+        ],
+    }
+
+
+def render_rubric(rubric: list[dict[str, Any]]) -> list[str]:
+    if not rubric:
+        return ["- No rubric items found."]
+    return [f"- `{item['id']}` ({item['weight']}): {item['description']}" for item in rubric]
+
+
+def render_markdown(payload: dict[str, Any]) -> str:
+    summary = payload["summary"]
+    contract = payload["review_contract"]
+    lines = [
+        "# Output Review Kit",
+        "",
+        "This reviewer-facing packet contains the blind A/B cases, decision fields, and command flow. It intentionally does not expose the answer key.",
+        "",
+        "## Summary",
+        "",
+        f"- cases: `{summary['case_count']}`",
+        f"- ready for adjudication: `{summary['ready_for_adjudication_count']}`",
+        f"- pending decisions: `{summary['pending_decision_count']}`",
+        f"- invalid decisions: `{summary['invalid_decision_count']}`",
+        f"- reviewer metadata present: `{str(summary['reviewer_metadata_present']).lower()}`",
+        f"- answer key hidden: `{str(summary['answer_key_hidden']).lower()}`",
+        f"- answer key path exposed: `{str(summary['answer_key_path_exposed']).lower()}`",
+        "",
+        "## Review Flow",
+        "",
+    ]
+    lines.extend(f"{index}. {step}" for index, step in enumerate(contract["reviewer_steps"], start=1))
+    lines.extend(["", "## Required Fields", ""])
+    lines.extend(f"- `{key}`: {value}" for key, value in contract["required_fields"].items())
+    lines.extend(["", "## Privacy Contract", ""])
+    lines.extend(f"- {item}" for item in contract["privacy_contract"])
+    lines.extend(
+        [
+            "",
+            "## Decision States",
+            "",
+            "| Case | State | Winner | Confidence | Reason | Blocking Reason |",
+            "| --- | --- | --- | --- | --- | --- |",
+        ]
+    )
+    for case in payload["cases"]:
+        state = case["decision_state"]
+        lines.append(
+            f"| `{case['case_id']}` | `{state['status']}` | `{str(state['winner_variant_recorded']).lower()}` | "
+            f"`{str(state['confidence_recorded']).lower()}` | `{str(state['reason_recorded']).lower()}` | {state['blocking_reason']} |"
+        )
+    lines.extend(["", "## Blind Cases", ""])
+    for case in payload["cases"]:
+        lines.extend(
+            [
+                f"### {case['case_id']}",
+                "",
+                f"Prompt: {case['prompt']}",
+                "",
+                "Rubric:",
+                *render_rubric(case["rubric"]),
+                "",
+                "#### Variant A",
+                "",
+                case["variant_a"]["output"] or "_No output found._",
+                "",
+                "#### Variant B",
+                "",
+                case["variant_b"]["output"] or "_No output found._",
+                "",
+            ]
+        )
+    if payload.get("failures"):
+        lines.extend(["## Failures", ""])
+        lines.extend(f"- {failure}" for failure in payload["failures"])
+    return "\n".join(lines).strip() + "\n"
+
+
+def prepare_output_review_kit(
+    blind_pack_json: Path,
+    blind_pack_md: Path,
+    decisions_path: Path,
+    output_json: Path,
+    output_md: Path,
+    write_template: bool = False,
+) -> dict[str, Any]:
+    blind_pack, failures = load_json(blind_pack_json)
+    decisions_payload, decision_failures = load_optional_decisions(decisions_path)
+    failures.extend(decision_failures)
+    template_written = False
+    if write_template and blind_pack and not decisions_path.exists():
+        decisions_path.parent.mkdir(parents=True, exist_ok=True)
+        decisions_path.write_text(
+            json.dumps(build_decision_template(blind_pack), ensure_ascii=False, indent=2) + "\n",
+            encoding="utf-8",
+        )
+        template_written = True
+        decisions_payload, decision_failures = load_optional_decisions(decisions_path)
+        failures.extend(decision_failures)
+    decisions, index_failures = decision_index(decisions_payload)
+    failures.extend(index_failures)
+    case_ids = {str(pair.get("case_id", "")) for pair in pairs_from_pack(blind_pack)}
+    for case_id in decisions:
+        if case_id not in case_ids:
+            failures.append(f"decision references unknown case_id: {case_id}")
+    cases = build_cases(blind_pack, decisions)
+    summary = build_summary(cases, failures, str(decisions_payload.get("reviewer", "")), str(decisions_payload.get("reviewed_at", "")))
+    payload = {
+        "schema_version": "1.0",
+        "ok": not failures,
+        "summary": summary,
+        "artifacts": {
+            "reviewer_kit_json": display_path(output_json),
+            "reviewer_kit_markdown": display_path(output_md),
+            "blind_pack_json": display_path(blind_pack_json),
+            "blind_pack_markdown": display_path(blind_pack_md),
+            "decisions": display_path(decisions_path),
+            "adjudication_json": "reports/output_review_adjudication.json",
+            "adjudication_markdown": "reports/output_review_adjudication.md",
+        },
+        "template_written": template_written,
+        "review_contract": build_contract(blind_pack_md, decisions_path),
+        "cases": cases,
+        "failures": failures,
+    }
+    output_json.parent.mkdir(parents=True, exist_ok=True)
+    output_md.parent.mkdir(parents=True, exist_ok=True)
+    output_json.write_text(json.dumps(payload, ensure_ascii=False, indent=2) + "\n", encoding="utf-8")
+    output_md.write_text(render_markdown(payload), encoding="utf-8")
+    return payload
+
+
+def main() -> None:
+    parser = argparse.ArgumentParser(description="Prepare a reviewer-facing blind A/B output review kit.")
+    parser.add_argument("--blind-pack-json", default=str(DEFAULT_BLIND_PACK_JSON))
+    parser.add_argument("--blind-pack-md", default=str(DEFAULT_BLIND_PACK_MD))
+    parser.add_argument("--decisions", default=str(DEFAULT_DECISIONS))
+    parser.add_argument("--output-json", default=str(DEFAULT_OUTPUT_JSON))
+    parser.add_argument("--output-md", default=str(DEFAULT_OUTPUT_MD))
+    parser.add_argument("--write-template", action="store_true")
+    args = parser.parse_args()
+
+    payload = prepare_output_review_kit(
+        Path(args.blind_pack_json).resolve(),
+        Path(args.blind_pack_md).resolve(),
+        Path(args.decisions).resolve(),
+        Path(args.output_json).resolve(),
+        Path(args.output_md).resolve(),
+        write_template=args.write_template,
+    )
+    print(json.dumps(payload, ensure_ascii=False, indent=2))
+    raise SystemExit(0 if payload["ok"] else 2)
+
+
+if __name__ == "__main__":
+    main()
@@ -41,8 +41,9 @@ TASK_TEMPLATES: dict[str, dict[str, Any]] = {
        "owner": "human reviewer",
        "objective": "Record real blind A/B reviewer decisions before claiming human output review completion.",
        "runbook": [
+            "python3 scripts/yao.py output-review-kit --write-template",
+            "Open reports/output_review_kit.md and choose A or B for each pair without opening the answer key.",
            "python3 scripts/adjudicate_output_review.py --write-template",
-            "Open reports/output_blind_review_pack.md and choose A or B for each pair without opening the answer key.",
            "Edit reports/output_review_decisions.json with winner_variant values and reviewer metadata.",
            "python3 scripts/yao.py output-review",
            "python3 scripts/yao.py skill-os2-audit . --generated-at <YYYY-MM-DD>",
@@ -55,6 +56,7 @@ TASK_TEMPLATES: dict[str, dict[str, Any]] = {
        ],
        "evidence_artifacts": [
            "reports/output_blind_review_pack.md",
+            "reports/output_review_kit.md",
            "reports/output_review_decisions.json",
            "reports/output_review_adjudication.json",
            "reports/output_review_adjudication.md",
@@ -63,6 +63,7 @@ def evidence_paths(skill_dir: Path) -> dict[str, str]:
        "output_eval": "reports/output_quality_scorecard.md",
        "output_execution": "reports/output_execution_runs.md",
        "output_blind_review": "reports/output_blind_review_pack.md",
+        "output_review_kit": "reports/output_review_kit.md",
        "output_review_decisions": "reports/output_review_decisions.json",
        "output_review_adjudication": "reports/output_review_adjudication.md",
        "benchmark_reproducibility": "reports/benchmark_reproducibility.md",
@@ -103,6 +104,7 @@ def load_review_data(skill_dir: Path) -> dict[str, dict[str, Any]]:
        "output_quality": load_json(reports / "output_quality_scorecard.json"),
        "output_execution": load_json(reports / "output_execution_runs.json"),
        "output_blind_review": load_json(reports / "output_blind_review_pack.json"),
+        "output_review_kit": load_json(reports / "output_review_kit.json"),
        "output_review_adjudication": load_json(reports / "output_review_adjudication.json"),
        "skill_os2_coverage": load_json(reports / "skill_os2_coverage.json"),
        "compiled_targets": load_json(reports / "compiled_targets.json"),
@@ -136,6 +138,7 @@ def insight_cards(data: dict[str, dict[str, Any]]) -> list[dict[str, str]]:
    output = data["output_quality"].get("summary", {})
    output_execution = data["output_execution"].get("summary", {})
    output_blind = data["output_blind_review"].get("summary", {})
+    output_review_kit = data["output_review_kit"].get("summary", {})
    output_review = data["output_review_adjudication"].get("summary", {})
    blueprint = data["skill_os2_coverage"].get("summary", {})
    compiled = data["compiled_targets"].get("summary", {})
@@ -184,6 +187,11 @@ def insight_cards(data: dict[str, dict[str, Any]]) -> list[dict[str, str]]:
            "value": str(output_blind.get("pair_count", 0)),
            "detail": "review pairs hide baseline vs with-skill labels",
        },
+        {
+            "label": "Review Kit",
+            "value": f"{output_review_kit.get('ready_for_adjudication_count', 0)}/{output_review_kit.get('case_count', 0)}",
+            "detail": f"pending {output_review_kit.get('pending_decision_count', 0)}; answer key hidden",
+        },
        {
            "label": "Review A/B",
            "value": f"{output_review.get('judgment_count', 0)}/{output_review.get('pair_count', 0)}",
@@ -78,7 +78,7 @@ def render_output_review_section(adjudication: dict[str, Any]) -> str:
    return (
        "<section>"
        "<h2>评审清单</h2>"
-        "<p class='muted'>每张卡片对应一个 blind A/B case，说明当前是否待判、答案是否仍隐藏、应填写的决策文件、有效字段和复跑命令。</p>"
+        "<p class='muted'>先打开 reports/output_review_kit.md；每张卡片对应一个 blind A/B case，说明当前是否待判、答案是否仍隐藏、应填写的决策文件、有效字段和复跑命令。</p>"
        f"{render_output_review_checklist(adjudication)}"
        "</section>"
    )
@@ -475,6 +475,7 @@ def build_report_model(skill_dir: Path) -> dict:
    output_quality = load_json(skill_dir / "reports" / "output_quality_scorecard.json")
    output_execution = load_json(skill_dir / "reports" / "output_execution_runs.json")
    output_blind_review = load_json(skill_dir / "reports" / "output_blind_review_pack.json")
+    output_review_kit = load_json(skill_dir / "reports" / "output_review_kit.json")
    output_review_adjudication = load_json(skill_dir / "reports" / "output_review_adjudication.json")
    benchmark_reproducibility = load_json(skill_dir / "reports" / "benchmark_reproducibility.json")
    conformance = load_json(skill_dir / "reports" / "conformance_matrix.json")
@@ -550,8 +551,11 @@ def build_report_model(skill_dir: Path) -> dict:
    if (skill_dir / "reports" / "output_blind_answer_key.json").exists():
        insert_after = deliverables.index("reports/output_blind_review_pack.md") + 1 if "reports/output_blind_review_pack.md" in deliverables else deliverables.index("reports/output_quality_scorecard.md") + 1
        deliverables.insert(insert_after, "reports/output_blind_answer_key.json")
+    if (skill_dir / "reports" / "output_review_kit.md").exists():
+        insert_after = deliverables.index("reports/output_blind_review_pack.md") + 1 if "reports/output_blind_review_pack.md" in deliverables else deliverables.index("reports/output_quality_scorecard.md") + 1
+        deliverables.insert(insert_after, "reports/output_review_kit.md")
    if (skill_dir / "reports" / "output_review_adjudication.md").exists():
-        insert_after = deliverables.index("reports/output_blind_answer_key.json") + 1 if "reports/output_blind_answer_key.json" in deliverables else deliverables.index("reports/output_quality_scorecard.md") + 1
+        insert_after = deliverables.index("reports/output_review_kit.md") + 1 if "reports/output_review_kit.md" in deliverables else deliverables.index("reports/output_quality_scorecard.md") + 1
        deliverables.insert(insert_after, "reports/output_review_adjudication.md")
    if (skill_dir / "reports" / "benchmark_reproducibility.md").exists():
        insert_after = deliverables.index("reports/output_review_adjudication.md") + 1 if "reports/output_review_adjudication.md" in deliverables else deliverables.index("reports/output_quality_scorecard.md") + 1
@@ -680,6 +684,12 @@ def build_report_model(skill_dir: Path) -> dict:
            "pair_count": output_blind_review.get("summary", {}).get("pair_count", 0),
            "answer_key_separate": output_blind_review.get("summary", {}).get("answer_key_separate", False),
        },
+        "output_review_kit": {
+            "ok": output_review_kit.get("ok", False),
+            "summary": output_review_kit.get("summary", {}),
+            "artifacts": output_review_kit.get("artifacts", {}),
+            "failures": output_review_kit.get("failures", []),
+        },
        "output_review_adjudication": {
            "ok": output_review_adjudication.get("ok", False),
            "summary": output_review_adjudication.get("summary", {}),
@@ -167,6 +167,7 @@ def command_report(args: argparse.Namespace) -> int:
            run_script("compile_skill.py", [str(ROOT)]),
            run_script("run_output_eval.py", []),
            run_script("run_output_execution.py", ["--runner-command", local_output_runner_command()]),
+            run_script("prepare_output_review_kit.py", []),
            run_script("adjudicate_output_review.py", []),
            run_script("render_adoption_drift_report.py", [str(ROOT)]),
            run_script("render_telemetry_hook_recipes.py", [str(ROOT)]),
@@ -201,6 +202,7 @@ def command_report(args: argparse.Namespace) -> int:
            "system_model": "reports/system-model.json",
            "compiled_targets": "reports/compiled_targets.json",
            "output_execution": "reports/output_execution_runs.json",
+            "output_review_kit": "reports/output_review_kit.json",
            "output_review_adjudication": "reports/output_review_adjudication.json",
            "adoption_drift": "reports/adoption_drift_report.json",
            "telemetry_hooks": "reports/telemetry_hook_recipes.json",
@@ -501,6 +503,25 @@ def command_output_execution(args: argparse.Namespace) -> int:
    return 0 if result["ok"] else 2


+def command_output_review_kit(args: argparse.Namespace) -> int:
+    cmd = []
+    if args.blind_pack_json:
+        cmd.extend(["--blind-pack-json", args.blind_pack_json])
+    if args.blind_pack_md:
+        cmd.extend(["--blind-pack-md", args.blind_pack_md])
+    if args.decisions:
+        cmd.extend(["--decisions", args.decisions])
+    if args.output_json:
+        cmd.extend(["--output-json", args.output_json])
+    if args.output_md:
+        cmd.extend(["--output-md", args.output_md])
+    if args.write_template:
+        cmd.append("--write-template")
+    result = run_script("prepare_output_review_kit.py", cmd)
+    print(json.dumps(result["payload"] if result["payload"] is not None else result, ensure_ascii=False, indent=2))
+    return 0 if result["ok"] else 2
+
+
 def command_output_review(args: argparse.Namespace) -> int:
    cmd = []
    if args.blind_pack:
@@ -587,6 +587,18 @@ def build_parser(command_handlers: dict[str, Callable[[argparse.Namespace], int]
    output_eval_cmd.add_argument("--blind-answer-key-json")
    output_eval_cmd.set_defaults(func=_handler(command_handlers, "command_output_eval"))

+    output_review_kit_cmd = subparsers.add_parser(
+        "output-review-kit",
+        help="Prepare a reviewer-facing blind A/B output review kit without exposing the answer key.",
+    )
+    output_review_kit_cmd.add_argument("--blind-pack-json")
+    output_review_kit_cmd.add_argument("--blind-pack-md")
+    output_review_kit_cmd.add_argument("--decisions")
+    output_review_kit_cmd.add_argument("--output-json")
+    output_review_kit_cmd.add_argument("--output-md")
+    output_review_kit_cmd.add_argument("--write-template", action="store_true")
+    output_review_kit_cmd.set_defaults(func=_handler(command_handlers, "command_output_review_kit"))
+
    output_execution_cmd = subparsers.add_parser(
        "output-exec",
        help="Record output-eval execution evidence, timing, and token usage.",
@@ -129,6 +129,7 @@
      "scripts/lint_skill.py",
      "scripts/local_output_eval_runner.py",
      "scripts/optimize_description.py",
+      "scripts/prepare_output_review_kit.py",
      "scripts/prepare_world_class_submission_kit.py",
      "scripts/probe_runtime_permissions.py",
      "scripts/promotion_checker.py",
@@ -186,8 +187,7 @@
      "scripts/skill_report_layout.py",
      "scripts/skill_report_metrics.py",
      "scripts/skill_report_model.py",
-      "scripts/sync_local_install.py",
-      "scripts/telemetry_native_host.py"
+      "scripts/sync_local_install.py"
    ],
    "assets": [
      "templates/basic_skill.md.j2",
@@ -41,6 +41,7 @@
        "scripts/lint_skill.py",
        "scripts/local_output_eval_runner.py",
        "scripts/optimize_description.py",
+        "scripts/prepare_output_review_kit.py",
        "scripts/prepare_world_class_submission_kit.py",
        "scripts/probe_runtime_permissions.py",
        "scripts/promotion_checker.py",
@@ -0,0 +1,203 @@
+#!/usr/bin/env python3
+import json
+import shutil
+import subprocess
+import sys
+from pathlib import Path
+
+
+ROOT = Path(__file__).resolve().parent.parent
+OUTPUT_EVAL = ROOT / "scripts" / "run_output_eval.py"
+REVIEW_KIT = ROOT / "scripts" / "prepare_output_review_kit.py"
+CLI = ROOT / "scripts" / "yao.py"
+
+
+def run(args: list[str], check: bool = True) -> subprocess.CompletedProcess[str]:
+    return subprocess.run(
+        [sys.executable, *args],
+        cwd=ROOT,
+        capture_output=True,
+        text=True,
+        check=check,
+    )
+
+
+def assert_no_answer_key_leak(text: str) -> None:
+    forbidden = [
+        "output_blind_answer_key",
+        "expected_winner",
+        "variant_a_role",
+        "variant_b_role",
+        "score_winner_role",
+    ]
+    for needle in forbidden:
+        assert needle not in text, f"review kit leaked hidden field {needle!r}"
+
+
+def main() -> None:
+    tmp_root = ROOT / "tests" / "tmp_output_review_kit"
+    if tmp_root.exists():
+        shutil.rmtree(tmp_root)
+    tmp_root.mkdir(parents=True, exist_ok=True)
+
+    scorecard_json = tmp_root / "output_quality_scorecard.json"
+    scorecard_md = tmp_root / "output_quality_scorecard.md"
+    blind_pack_json = tmp_root / "output_blind_review_pack.json"
+    blind_pack_md = tmp_root / "output_blind_review_pack.md"
+    answer_key_json = tmp_root / "output_blind_answer_key.json"
+
+    run(
+        [
+            str(OUTPUT_EVAL),
+            "--cases",
+            str(ROOT / "evals" / "output" / "cases.jsonl"),
+            "--output-json",
+            str(scorecard_json),
+            "--output-md",
+            str(scorecard_md),
+            "--blind-pack-json",
+            str(blind_pack_json),
+            "--blind-pack-md",
+            str(blind_pack_md),
+            "--blind-answer-key-json",
+            str(answer_key_json),
+        ]
+    )
+
+    decisions_path = tmp_root / "output_review_decisions.json"
+    kit_json = tmp_root / "output_review_kit.json"
+    kit_md = tmp_root / "output_review_kit.md"
+    pending_proc = run(
+        [
+            str(REVIEW_KIT),
+            "--blind-pack-json",
+            str(blind_pack_json),
+            "--blind-pack-md",
+            str(blind_pack_md),
+            "--decisions",
+            str(decisions_path),
+            "--output-json",
+            str(kit_json),
+            "--output-md",
+            str(kit_md),
+            "--write-template",
+        ]
+    )
+    pending_payload = json.loads(pending_proc.stdout)
+    assert pending_payload["ok"], pending_payload
+    assert pending_payload["template_written"], pending_payload
+    assert pending_payload["summary"]["case_count"] == 5, pending_payload
+    assert pending_payload["summary"]["pending_decision_count"] == 5, pending_payload
+    assert pending_payload["summary"]["ready_for_adjudication_count"] == 0, pending_payload
+    assert pending_payload["summary"]["answer_key_hidden"] is True, pending_payload
+    assert pending_payload["summary"]["answer_key_path_exposed"] is False, pending_payload
+    assert pending_payload["summary"]["ready_to_run_adjudication"] is False, pending_payload
+    assert "answer_key" not in pending_payload["artifacts"], pending_payload["artifacts"]
+    assert len(pending_payload["cases"]) == 5, pending_payload["cases"]
+    first_case = pending_payload["cases"][0]
+    assert first_case["variant_a"]["output"], first_case
+    assert first_case["variant_b"]["output"], first_case
+    assert first_case["decision_state"]["status"] == "awaiting-decision", first_case
+    assert decisions_path.exists(), decisions_path
+
+    kit_json_text = kit_json.read_text(encoding="utf-8")
+    kit_md_text = kit_md.read_text(encoding="utf-8")
+    assert "# Output Review Kit" in kit_md_text, kit_md_text
+    assert "answer key path exposed: `false`" in kit_md_text, kit_md_text
+    assert "Variant A" in kit_md_text and "Variant B" in kit_md_text, kit_md_text
+    assert "python3 scripts/yao.py output-review" in kit_md_text, kit_md_text
+    assert_no_answer_key_leak(kit_json_text)
+    assert_no_answer_key_leak(kit_md_text)
+
+    cli_proc = run(
+        [
+            str(CLI),
+            "output-review-kit",
+            "--blind-pack-json",
+            str(blind_pack_json),
+            "--blind-pack-md",
+            str(blind_pack_md),
+            "--decisions",
+            str(decisions_path),
+            "--output-json",
+            str(tmp_root / "cli_output_review_kit.json"),
+            "--output-md",
+            str(tmp_root / "cli_output_review_kit.md"),
+        ]
+    )
+    cli_payload = json.loads(cli_proc.stdout)
+    assert cli_payload["ok"], cli_payload
+    assert cli_payload["summary"]["case_count"] == 5, cli_payload
+    assert cli_payload["summary"]["answer_key_hidden"] is True, cli_payload
+    assert "answer_key" not in cli_payload["artifacts"], cli_payload["artifacts"]
+
+    template = json.loads(decisions_path.read_text(encoding="utf-8"))
+    filled = {
+        "schema_version": "1.0",
+        "reviewer": "Yao QA",
+        "reviewed_at": "2026-06-14",
+        "decisions": [
+            {
+                "case_id": item["case_id"],
+                "winner_variant": "A",
+                "confidence": 0.8,
+                "reason": "Variant A is clearer against the visible rubric.",
+            }
+            for item in template["decisions"]
+        ],
+    }
+    decisions_path.write_text(json.dumps(filled, ensure_ascii=False, indent=2) + "\n", encoding="utf-8")
+    ready_proc = run(
+        [
+            str(REVIEW_KIT),
+            "--blind-pack-json",
+            str(blind_pack_json),
+            "--blind-pack-md",
+            str(blind_pack_md),
+            "--decisions",
+            str(decisions_path),
+            "--output-json",
+            str(tmp_root / "ready_output_review_kit.json"),
+            "--output-md",
+            str(tmp_root / "ready_output_review_kit.md"),
+        ]
+    )
+    ready_payload = json.loads(ready_proc.stdout)
+    assert ready_payload["ok"], ready_payload
+    assert ready_payload["summary"]["ready_for_adjudication_count"] == 5, ready_payload
+    assert ready_payload["summary"]["pending_decision_count"] == 0, ready_payload
+    assert ready_payload["summary"]["invalid_decision_count"] == 0, ready_payload
+    assert ready_payload["summary"]["reviewer_metadata_present"] is True, ready_payload
+    assert ready_payload["summary"]["ready_to_run_adjudication"] is True, ready_payload
+    assert all(case["decision_state"]["status"] == "ready-for-adjudication" for case in ready_payload["cases"]), ready_payload
+
+    invalid = dict(filled)
+    invalid["decisions"] = [dict(filled["decisions"][0], winner_variant="C", reason="")]
+    invalid_path = tmp_root / "invalid_output_review_decisions.json"
+    invalid_path.write_text(json.dumps(invalid, ensure_ascii=False, indent=2) + "\n", encoding="utf-8")
+    invalid_proc = run(
+        [
+            str(REVIEW_KIT),
+            "--blind-pack-json",
+            str(blind_pack_json),
+            "--blind-pack-md",
+            str(blind_pack_md),
+            "--decisions",
+            str(invalid_path),
+            "--output-json",
+            str(tmp_root / "invalid_output_review_kit.json"),
+            "--output-md",
+            str(tmp_root / "invalid_output_review_kit.md"),
+        ]
+    )
+    invalid_payload = json.loads(invalid_proc.stdout)
+    assert invalid_payload["ok"], invalid_payload
+    assert invalid_payload["summary"]["invalid_decision_count"] == 1, invalid_payload
+    assert invalid_payload["summary"]["answer_key_hidden"] is True, invalid_payload
+    assert invalid_payload["cases"][0]["decision_state"]["status"] == "needs-fix", invalid_payload["cases"][0]
+
+    print(json.dumps({"ok": True}, ensure_ascii=False, indent=2))
+
+
+if __name__ == "__main__":
+    main()
@@ -89,6 +89,7 @@ def main() -> None:
    assert "world-class-submission-review" in parser_help, parser_help
    assert "world-class-claim-guard" in parser_help, parser_help
    assert "benchmark-reproducibility" in parser_help, parser_help
+    assert "output-review-kit" in parser_help, parser_help
    assert "telemetry-import" in parser_help, parser_help
    assert "telemetry-emit" in parser_help, parser_help
    assert "telemetry-hooks" in parser_help, parser_help
@@ -757,6 +758,7 @@ def main() -> None:
    assert "prompt_quality_profile" in report_result["payload"]["artifacts"], report_result
    assert "compiled_targets" in report_result["payload"]["artifacts"], report_result
    assert "output_execution" in report_result["payload"]["artifacts"], report_result
+    assert "output_review_kit" in report_result["payload"]["artifacts"], report_result
    assert "output_review_adjudication" in report_result["payload"]["artifacts"], report_result
    assert "adoption_drift" in report_result["payload"]["artifacts"], report_result
    assert "review_annotations" in report_result["payload"]["artifacts"], report_result