diff --git a/reports/context_budget.json b/reports/context_budget.json index 271a99b0..9b92a277 100644 --- a/reports/context_budget.json +++ b/reports/context_budget.json @@ -6,15 +6,15 @@ "context_budget_tier": "production", "context_budget_limit": 1000, "skill_body_tokens": 797, - "other_text_tokens": 1070383, + "other_text_tokens": 1070810, "estimated_initial_load_tokens": 990, - "estimated_total_text_tokens": 1071180, - "deferred_resource_tokens": 495462, + "estimated_total_text_tokens": 1071607, + "deferred_resource_tokens": 495533, "deferred_resource_warn_threshold": 120000, "deferred_resource_dirs": [ { "path": "scripts", - "estimated_tokens": 435364, + "estimated_tokens": 435435, "file_count": 139 }, { @@ -36,7 +36,7 @@ "large_deferred_resource_dirs": [ { "path": "scripts", - "estimated_tokens": 435364, + "estimated_tokens": 435435, "file_count": 139 } ], @@ -59,7 +59,7 @@ ], "missing": [], "path": "scripts", - "estimated_tokens": 435364, + "estimated_tokens": 435435, "file_count": 139, "rationale": "Script resources are deterministic deferred tools, not initial-load prompt context." } diff --git a/reports/context_budget.md b/reports/context_budget.md index 8e3ba33a..fa5c305f 100644 --- a/reports/context_budget.md +++ b/reports/context_budget.md @@ -2,7 +2,7 @@ | Target | Path | Tier | Limit | Initial | SKILL | Deferred | Resource Governance | Large Deferred Dirs | Quality Density | Unused Dirs | Status | | --- | --- | --- | ---: | ---: | ---: | ---: | --- | --- | ---: | --- | --- | -| root | `.` | `production` | 1000 | 990 | 797 | 495462 | `governed` | scripts:435364 | 131.3 | - | ok | +| root | `.` | `production` | 1000 | 990 | 797 | 495533 | `governed` | scripts:435435 | 131.3 | - | ok | | complex-release-orchestrator | `examples/complex-release-orchestrator/generated-skill` | `production` | 1000 | 790 | 718 | 1657 | `not-required` | - | 164.6 | - | ok | | governed-incident-command | `examples/governed-incident-command/generated-skill` | `production` | 1000 | 760 | 658 | 1030 | `not-required` | - | 171.1 | - | ok | diff --git a/reports/context_budget_summary.json b/reports/context_budget_summary.json index aba4d654..8dc31584 100644 --- a/reports/context_budget_summary.json +++ b/reports/context_budget_summary.json @@ -8,11 +8,11 @@ "budget_limit": 1000, "initial_tokens": 990, "skill_body_tokens": 797, - "deferred_resource_tokens": 495462, + "deferred_resource_tokens": 495533, "large_deferred_resource_dirs": [ { "path": "scripts", - "estimated_tokens": 435364, + "estimated_tokens": 435435, "file_count": 139 } ], @@ -35,7 +35,7 @@ ], "missing": [], "path": "scripts", - "estimated_tokens": 435364, + "estimated_tokens": 435435, "file_count": 139, "rationale": "Script resources are deterministic deferred tools, not initial-load prompt context." } diff --git a/reports/security_trust_report.json b/reports/security_trust_report.json index fd4b2e61..61f00cf3 100644 --- a/reports/security_trust_report.json +++ b/reports/security_trust_report.json @@ -23,7 +23,7 @@ "interactive_script_count": 0, "package_hash_scope": "source-contract-without-generated-reports", "package_hash_file_count": 230, - "package_sha256": "2bd5ae0684cf7508cce7611a6a840c287d35067f13e5f78f33c45667c501280c" + "package_sha256": "81880e6374f9824422252d99c7b657d49eb626c2bd8ac2f9710317fb977417a8" }, "failures": [], "warnings": [], diff --git a/reports/security_trust_report.md b/reports/security_trust_report.md index 56f33c13..35dc1e53 100644 --- a/reports/security_trust_report.md +++ b/reports/security_trust_report.md @@ -16,7 +16,7 @@ - Interactive scripts: `0` - Package hash scope: `source-contract-without-generated-reports` - Package hash files: `230` -- Package SHA256: `2bd5ae0684cf7508cce7611a6a840c287d35067f13e5f78f33c45667c501280c` +- Package SHA256: `81880e6374f9824422252d99c7b657d49eb626c2bd8ac2f9710317fb977417a8` ## Failures diff --git a/reports/skill-os-2-review.md b/reports/skill-os-2-review.md index 9591ef6d..09788f1e 100644 --- a/reports/skill-os-2-review.md +++ b/reports/skill-os-2-review.md @@ -52,7 +52,7 @@ Yao Meta Skill is no longer only a Meta Skill factory. The current working tree - Review Studio now avoids over-claiming release readiness when blind A/B adjudication, waiver handling, and world-class evidence are still pending: the root Meta Skill is in `review` with score `91`, no blockers, three warnings, and explicit actions for Output Lab reviewer adjudication, waiver handling, and world-class evidence completion. - Review Studio Output Lab actions now link directly to `reports/output_review_decisions.json`, so pending blind A/B reviewer decisions have a concrete template instead of only a general adjudication warning. - Output Review Adjudication now preserves blind-review integrity by hiding expected winners for pending or invalid reviewer decisions; answer keys are revealed only after a valid A/B decision exists for that case. -- Output Review Import v0 now accepts reviewer JSON, JSONL, or CSV decision sources, recursively rejects raw prompt/output/transcript/message and answer-key fields, writes canonical `reports/output_review_decisions.json`, and can run adjudication immediately without opening the answer key before valid decisions exist. +- Output Review Import v0 now accepts reviewer JSON, JSONL, or CSV decision sources, recursively rejects raw prompt/output/transcript/message, credential, secret, token, and answer-key fields case-insensitively, writes canonical `reports/output_review_decisions.json`, and can run adjudication immediately without opening the answer key before valid decisions exist. - Provider Output Eval Runner v0 so `python3 scripts/yao.py output-exec --provider-runner openai` can collect real provider-backed model evidence through a reviewed OpenAI Responses API compatible runner instead of ad hoc shell glue. - Weekly SkillOps Curator v0 so daily SkillOps opportunities, Skill Atlas portfolio signals, release-lock state, and world-class evidence gaps roll up into a proposal-only weekly maintenance queue without source-file writes. diff --git a/scripts/import_output_review_decisions.py b/scripts/import_output_review_decisions.py index 89430766..de45d10a 100644 --- a/scripts/import_output_review_decisions.py +++ b/scripts/import_output_review_decisions.py @@ -24,6 +24,7 @@ SCRIPT_INTERFACE = "cli" SCRIPT_INTERFACE_REASON = "Imports human blind A/B reviewer decisions into the canonical output-review decision file." RAW_CONTENT_FIELDS = { + "api_key", "prompt", "prompts", "input", @@ -35,18 +36,28 @@ RAW_CONTENT_FIELDS = { "message", "messages", "user_message", + "user_messages", "assistant_message", + "assistant_messages", "model_output", "baseline_output", "with_skill_output", "raw_content", "raw_prompt", + "raw_provider_prompt", + "raw_user_content", "raw_output", + "credential", + "credentials", + "secret", + "secrets", + "token", } ANSWER_KEY_FIELDS = { "expected", "expected_winner", + "expected_winner_role", "expected_winner_variant", "answer_key", "label", @@ -148,7 +159,8 @@ def forbidden_field_paths(value: Any, prefix: str) -> list[str]: for key, child in value.items(): key_text = str(key).strip() child_path = f"{prefix}.{key_text}" - if key_text in RAW_CONTENT_FIELDS or key_text in ANSWER_KEY_FIELDS: + normalized_key = key_text.casefold() + if normalized_key in RAW_CONTENT_FIELDS or normalized_key in ANSWER_KEY_FIELDS: found.append(child_path) found.extend(forbidden_field_paths(child, child_path)) elif isinstance(value, list): diff --git a/tests/verify_output_review_adjudication.py b/tests/verify_output_review_adjudication.py index f6afcd25..615a448d 100644 --- a/tests/verify_output_review_adjudication.py +++ b/tests/verify_output_review_adjudication.py @@ -324,7 +324,7 @@ def main() -> None: { "case_id": answer_key["answers"][0]["case_id"], "winner_variant": "A", - "metadata": {"raw_output": "nested raw output must not be imported"}, + "metadata": {"Raw_Output": "nested raw output must not be imported"}, } ], }, @@ -349,9 +349,47 @@ def main() -> None: nested_private_payload = json.loads(nested_private_proc.stdout) assert nested_private_proc.returncode == 2, nested_private_payload assert nested_private_payload["ok"] is False, nested_private_payload - assert any("decision #1.metadata.raw_output" in failure for failure in nested_private_payload["failures"]), nested_private_payload + assert any("decision #1.metadata.Raw_Output" in failure for failure in nested_private_payload["failures"]), nested_private_payload assert not (tmp_root / "nested_private_decisions.json").exists(), nested_private_payload + credential_source = tmp_root / "credential_source.json" + credential_source.write_text( + json.dumps( + { + "reviewer": "Yao QA", + "reviewed_at": "2026-06-13", + "decisions": [ + { + "case_id": answer_key["answers"][0]["case_id"], + "winner_variant": "A", + "metadata": {"API_KEY": "credential material must not be imported"}, + } + ], + }, + ensure_ascii=False, + indent=2, + ) + + "\n", + encoding="utf-8", + ) + credential_proc = run( + [ + str(IMPORTER), + "--input", + str(credential_source), + "--blind-pack", + str(blind_pack_json), + "--output-json", + str(tmp_root / "credential_decisions.json"), + ], + check=False, + ) + credential_payload = json.loads(credential_proc.stdout) + assert credential_proc.returncode == 2, credential_payload + assert credential_payload["ok"] is False, credential_payload + assert any("decision #1.metadata.API_KEY" in failure for failure in credential_payload["failures"]), credential_payload + assert not (tmp_root / "credential_decisions.json").exists(), credential_payload + unknown_case = tmp_root / "unknown_case.json" unknown_case.write_text( json.dumps(