vignette/docs/ops/evidence/continuous-improvement-agentic-producer-live-2026-08-07.json
Yun Chan 16e791e044 G0~G8 성과·동맹 측정 OS 작업 일괄 고정
8월 7일까지 워킹트리에만 남아 있던 미커밋 작업을 커밋한다. 여러 사본
폴더(worktree·clone)에 흩어져 있던 중간 스냅샷을 정리하기 전에 원본을
git 이력으로 고정하는 것이 목적이다.

- contracts/routes/services: measurement, outcome_trajectory, rupture_repair,
  deliberate_practice, calibration_transfer, supervision_research,
  multimodal_alliance, continuous_improvement 계열 신규 모듈과 테스트
- infra/db/init: 07~16 마이그레이션(측정 기반~calibration transfer 실행)
- apps/web: 세션 리뷰 카드·관리 화면·E2E 스펙 추가
- docs/ops: G0~G8 라이브 통합·배포·롤백 증거 문서와 evidence JSON/PNG
- scripts: smoke·ledger·릴리스 에이전트·NAS 프리뷰 운영 스크립트

engine.public 로그 .bak과 apps/web/test-results 산출물은 커밋에서 제외했다.
2026-08-08 01:30:53 +09:00

113 lines
3.3 KiB
JSON

{
"evidence_version": "g8-agentic-producer-live-v1",
"environment": "dev",
"engine_ready": {
"ok": true,
"detail": "OK"
},
"repo_source": {
"job_key": "oas-g8-job-repo-synthetic-rupture-v5",
"data_classification": "synthetic_replay_red_team_coverage_drift",
"source_ids": [
"oas-g8-source-repo-synthetic-rupture-v5"
],
"usage_statuses": [
"approved"
],
"content_hashes_verified": true,
"raw_transcript_included": false,
"pii_included": false
},
"first_run": {
"status": "completed",
"engine_call_count": 7,
"stages": [
"benchmark_judge_1",
"benchmark_judge_2",
"benchmark_judge_3",
"benchmark_variant_generation",
"draft_generation",
"red_team_leakage-bias",
"red_team_safety-identity"
],
"provider_models": [
{
"provider": "claude_cli",
"model": "claude-opus-4-8"
},
{
"provider": "claude_cli",
"model": "claude-opus-4-8"
},
{
"provider": "claude_cli",
"model": "claude-opus-4-8"
},
{
"provider": "claude_cli",
"model": "claude-opus-4-8"
},
{
"provider": "claude_cli",
"model": "claude-opus-4-8"
},
{
"provider": "claude_cli",
"model": "claude-opus-4-8"
},
{
"provider": "claude_cli",
"model": "claude-opus-4-8"
}
],
"error_code": null
},
"durable_db": {
"job_id": "beb47b8c-40a3-5499-8716-84148af4308e",
"job_key": "oas-g8-job-repo-synthetic-rupture-v5",
"status": "completed",
"attempt_count": 1,
"data_classification": "synthetic_replay_red_team_coverage_drift",
"source_fingerprint": "d3dd9f2ad157d27f0738230eb1fa9cccba679bb0182d97498e637a4a2949df88",
"last_error_code": null,
"last_error_message": null,
"result_submission_id": "50aff622-fc77-5ede-9e83-eaaf8250ca02",
"result_qualification_id": "2713aa69-3f54-5749-b1f9-dda75d43c896",
"pipeline_id": "4a21eb74-5905-5a1e-a452-8e2ed348faeb",
"generation_model": "claude_cli/claude-opus-4-8",
"prompt_version": "1.4.0",
"prompt_sha256": "93d4ff3df5094f70619da3a944f1bcfbb5e2e5bc7b84707c2ab7602b29cfab66",
"payload_sha256": "99519444e640951660268f9855c6e36156dc82bd88577b793abfa8f5949b7b18",
"pipeline_state": "pending_human_approval",
"clinical_claim_allowed": false,
"gate_state": "pending_human_approval",
"catalog_entry_id": "oas-g8-catalog-4a21eb7459055a1ea4528e2ed348faeb",
"variant_count": 3,
"variant_pass_rate": 1.0,
"qualified": true,
"reviewer_count": 2,
"catalog_count": 0,
"approval_count": 0,
"sources": [
{
"version": "1.4.0",
"source_id": "oas-g8-source-repo-synthetic-rupture-v5",
"usage_status": "approved",
"content_sha256": "706076b5ac6ab3259e000f65087dcc29c939e6fe82b14d7210923a3c51a30651",
"provenance_uri": "repo://apps/api/app/data/continuous_improvement/synthetic_source_pack.v5.json"
}
]
},
"safety_boundary": {
"human_approval_required": true,
"catalog_promoted": false,
"human_approval_event_count": 0,
"clinical_claim_allowed": false
},
"idempotent_replay": {
"reported": true,
"engine_call_count": 0,
"same_submission_id": true,
"same_qualification_id": true
}
}