warning 可以被有边界地接受，但必须写入 reviewer、理由、范围和到期时间；blocker 与 world-class 完成证据不能通过 waiver 变成通过。
diff --git a/scripts/review_studio_data.py b/scripts/review_studio_data.py index dd56101..571aac2 100644 --- a/scripts/review_studio_data.py +++ b/scripts/review_studio_data.py @@ -29,6 +29,33 @@ def load_json(path: Path) -> dict[str, Any]: return payload if isinstance(payload, dict) else {} +def latest_json_report(directory: Path) -> dict[str, Any]: + if not directory.exists(): + return {} + candidates = sorted(path for path in directory.glob("*.json") if path.is_file()) + if not candidates: + return {} + payload = load_json(candidates[-1]) + if payload: + try: + payload.setdefault("source_path", str(candidates[-1].resolve().relative_to(directory.parent.parent.parent.resolve()))) + except ValueError: + payload.setdefault("source_path", str(candidates[-1])) + return payload + + +def latest_report_path(skill_dir: Path, directory: Path, suffix: str) -> str: + if not directory.exists(): + return "" + candidates = sorted(path for path in directory.glob(f"*{suffix}") if path.is_file()) + if not candidates: + return "" + try: + return str(candidates[-1].resolve().relative_to(skill_dir.resolve())) + except ValueError: + return str(candidates[-1]) + + def load_yaml(path: Path) -> dict[str, Any]: if not path.exists() or yaml is None: return {} @@ -107,6 +134,12 @@ def evidence_paths(skill_dir: Path) -> dict[str, str]: skill_ir_path = find_skill_ir_path(skill_dir) if skill_ir_path: rels["skill_ir"] = skill_ir_path + daily_md = latest_report_path(skill_dir, skill_dir / "reports" / "skillops" / "daily", ".md") + weekly_md = latest_report_path(skill_dir, skill_dir / "reports" / "skillops" / "weekly", ".md") + if daily_md: + rels["daily_skillops"] = daily_md + if weekly_md: + rels["weekly_curator"] = weekly_md return {key: rel for key, rel in rels.items() if (skill_dir / rel).exists() or (ROOT / rel).exists()} @@ -134,6 +167,8 @@ def load_review_data(skill_dir: Path) -> dict[str, dict[str, Any]]: "promotion": load_json(reports / "promotion_decisions.json"), "atlas": load_json(reports / "skill_atlas.json"), "adoption_drift": load_json(reports / "adoption_drift_report.json"), + "daily_skillops": latest_json_report(reports / "skillops" / "daily"), + "weekly_curator": latest_json_report(reports / "skillops" / "weekly"), "review_waivers": load_json(reports / "review_waivers.json"), "review_annotations": load_json(reports / "review_annotations.json"), "adaptation_proposals": load_json(reports / "adaptation_proposals.json"), @@ -173,6 +208,8 @@ def insight_cards(data: dict[str, dict[str, Any]]) -> list[dict[str, str]]: architecture = data["architecture_maintainability"].get("summary", {}) atlas = data["atlas"].get("summary", {}) adoption = data["adoption_drift"].get("summary", {}) + daily_skillops = data["daily_skillops"].get("summary", {}) + weekly_curator = data["weekly_curator"].get("summary", {}) waivers = data["review_waivers"].get("summary", {}) annotations = data["review_annotations"].get("summary", {}) intake = data["world_class_evidence_intake"].get("summary", {}) @@ -277,6 +314,24 @@ def insight_cards(data: dict[str, dict[str, Any]]) -> list[dict[str, str]]: "value": str(adoption.get("risk_band", "n/a")), "detail": f"{adoption.get('event_count', 0)} metadata events; {adoption.get('missed_trigger_count', 0)} missed triggers", }, + { + "label": "Daily Ops", + "value": str(daily_skillops.get("proposal_count", 0)), + "detail": ( + f"{daily_skillops.get('decision', 'missing')}; " + f"approval {daily_skillops.get('approval_count', 0)}; " + f"release lock {str(daily_skillops.get('release_lock_ready', False)).lower()}" + ), + }, + { + "label": "Weekly Queue", + "value": str(weekly_curator.get("unique_opportunity_count", 0)), + "detail": ( + f"{weekly_curator.get('decision', 'missing')}; " + f"ready {weekly_curator.get('ready_for_approval_review_count', 0)}; " + f"top score {weekly_curator.get('top_score', 0)}" + ), + }, { "label": "Waivers", "value": str(waivers.get("active_count", 0)), diff --git a/scripts/review_studio_formatting.py b/scripts/review_studio_formatting.py index 42db70e..e9107b3 100644 --- a/scripts/review_studio_formatting.py +++ b/scripts/review_studio_formatting.py @@ -27,7 +27,9 @@ LABELS = { "command_executed_count": "命令执行", "compatibility_pass_count": "兼容通过", "covered_gate_count": "覆盖 Gate", + "daily_report_count": "日报数", "declared_bump": "声明版本", + "decision": "决策", "delta": "增益", "event_count": "事件数", "failure_count": "失败数", @@ -73,8 +75,11 @@ LABELS = { "public_claim_blocker_count": "声明阻断", "public_claim_ready": "可公开声明", "public_world_class_ready": "世界级", + "proposal_count": "提案数", + "proposal_review_count": "提案复核", "reproducibility_ready": "本地复现", "recorded_fixture_count": "记录样本", + "ready_for_approval_review_count": "待批准复核", "ready_for_external_collection": "可收证据", "ready_for_ledger_review": "可审台账", "recommended_pr_count": "建议 PR", @@ -98,6 +103,7 @@ LABELS = { "token_estimated_count": "估算 Token", "token_observed_count": "真实 Token", "trust_level": "信任级别", + "unique_opportunity_count": "唯一机会", "variant_run_count": "运行数", "valid_submission_count": "有效提交", "version": "版本", diff --git a/scripts/review_studio_gates.py b/scripts/review_studio_gates.py index 57ea03f..60b7acd 100644 --- a/scripts/review_studio_gates.py +++ b/scripts/review_studio_gates.py @@ -438,6 +438,10 @@ def build_gates(skill_dir: Path, output_html: Path, data: dict[str, dict[str, An adoption = data["adoption_drift"] adoption_summary = adoption.get("summary", {}) + daily_skillops = data.get("daily_skillops", {}) + weekly_curator = data.get("weekly_curator", {}) + daily_summary = daily_skillops.get("summary", {}) if isinstance(daily_skillops, dict) else {} + weekly_summary = weekly_curator.get("summary", {}) if isinstance(weekly_curator, dict) else {} if not adoption: adoption_status = "warn" adoption_detail = "adoption drift report is missing; real usage impact is unknown" @@ -454,13 +458,43 @@ def build_gates(skill_dir: Path, output_html: Path, data: dict[str, dict[str, An f"bad-output {adoption_summary.get('bad_output_count', 0)}; " f"risk {risk_band}" ) + skillops_parts = [] + skillops_blocked = False + for label, report, summary in [ + ("daily", daily_skillops, daily_summary), + ("weekly", weekly_curator, weekly_summary), + ]: + if not report: + continue + writes_source = bool(summary.get("writes_source_files") or report.get("writes_source_files")) + auto_patch = bool(summary.get("auto_patch_enabled") or report.get("auto_patch_enabled")) + failure_count = int(summary.get("failure_count", 0) or 0) + if writes_source or auto_patch or failure_count: + skillops_blocked = True + if label == "daily": + skillops_parts.append( + f"daily proposals {summary.get('proposal_count', 0)}; " + f"daily decision {summary.get('decision', report.get('decision', 'unknown'))}; " + f"daily release lock {str(summary.get('release_lock_ready', False)).lower()}" + ) + else: + skillops_parts.append( + f"weekly queue {summary.get('unique_opportunity_count', 0)} unique; " + f"weekly ready {summary.get('ready_for_approval_review_count', 0)}; " + f"weekly top {summary.get('top_score', 0)}; " + f"weekly release lock {str(summary.get('release_lock_ready', False)).lower()}" + ) + if skillops_blocked: + adoption_status = "block" + if skillops_parts: + adoption_detail = adoption_detail + "; " + "; ".join(skillops_parts) gates.append( gate( "operations-loop", "运营回路", adoption_status, adoption_detail, - "reports/adoption_drift_report.json", + "reports/adoption_drift_report.json + reports/skillops/daily + reports/skillops/weekly", _report_link(output_html, skill_dir, "reports/adoption_drift_report.md"), ) ) diff --git a/tests/verify_review_studio.py b/tests/verify_review_studio.py index 5fd0eb7..065c8d7 100644 --- a/tests/verify_review_studio.py +++ b/tests/verify_review_studio.py @@ -370,7 +370,11 @@ def main() -> None: assert operations_gate["status"] == "pass", operations_gate assert "metadata events" in operations_gate["detail"], operations_gate assert "risk low" in operations_gate["detail"], operations_gate + assert "daily proposals" in operations_gate["detail"], operations_gate + assert "weekly queue" in operations_gate["detail"], operations_gate assert "reports/adoption_drift_report.json" in operations_gate["evidence"], operations_gate + assert "reports/skillops/daily" in operations_gate["evidence"], operations_gate + assert "reports/skillops/weekly" in operations_gate["evidence"], operations_gate waivers_gate = next(item for item in payload["gates"] if item["key"] == "review-waivers") assert waivers_gate["status"] == "warn", waivers_gate assert "1 warning gates still need reviewer decision" in waivers_gate["detail"], waivers_gate @@ -408,6 +412,8 @@ def main() -> None: assert full_payload["evidence_paths"]["benchmark_reproducibility"] == "reports/benchmark_reproducibility.md", full_payload["evidence_paths"] if (ROOT / "reports" / "skill_os2_coverage.md").exists(): assert full_payload["evidence_paths"]["skill_os2_coverage"] == "reports/skill_os2_coverage.md", full_payload["evidence_paths"] + assert full_payload["evidence_paths"]["daily_skillops"].startswith("reports/skillops/daily/"), full_payload["evidence_paths"] + assert full_payload["evidence_paths"]["weekly_curator"].startswith("reports/skillops/weekly/"), full_payload["evidence_paths"] assert full_payload["evidence_paths"]["review_annotations"] == "reports/review_annotations.md", full_payload["evidence_paths"] assert full_payload["evidence_paths"]["adaptation_proposals"] == "reports/adaptation_proposals.md", full_payload["evidence_paths"] assert full_payload["evidence_paths"]["adaptation_approval_ledger"] == "reports/adaptation_approval_ledger.json", full_payload["evidence_paths"] @@ -451,6 +457,13 @@ def main() -> None: assert all(not item["answer_key_visible"] for item in output_review_checklist), output_review_checklist assert output_review_checklist[0]["commands"]["adjudicate"] == "python3 scripts/yao.py output-review", output_review_checklist[0] assert full_payload["data"]["review_annotations"]["summary"]["annotation_count"] == 0, full_payload["data"]["review_annotations"] + daily_skillops_summary = full_payload["data"]["daily_skillops"]["summary"] + assert daily_skillops_summary["writes_source_files"] is False, daily_skillops_summary + assert daily_skillops_summary["auto_patch_enabled"] is False, daily_skillops_summary + weekly_curator_summary = full_payload["data"]["weekly_curator"]["summary"] + assert weekly_curator_summary["unique_opportunity_count"] >= 1, weekly_curator_summary + assert weekly_curator_summary["writes_source_files"] is False, weekly_curator_summary + assert weekly_curator_summary["auto_patch_enabled"] is False, weekly_curator_summary assert full_payload["data"]["adaptation_proposals"]["report_contract"]["contract"] == "adaptation-proposals", full_payload["data"]["adaptation_proposals"] assert full_payload["data"]["adaptation_proposals"]["proposal_count"] == full_payload["data"]["adaptation_proposals"]["summary"]["proposal_count"], full_payload["data"]["adaptation_proposals"] assert full_payload["data"]["adaptation_approval_ledger"]["report_contract"]["contract"] == "adaptation-approval-ledger", full_payload["data"]["adaptation_approval_ledger"] @@ -651,6 +664,9 @@ def main() -> None: assert "reports/world_class_evidence_intake.md" in html, html assert "reports/world_class_operator_runbook.html" in html, html assert "reports/world_class_claim_guard.md" in html, html + assert "日常运维" in html, html + assert "周度队列" in html, html + assert "Weekly Queue" in html, html assert "英文完成断言、true 状态声明或中文完成态" in html, html assert "world-evidence-grid" in html, html assert "Provider Holdout" in html, html