Antoine/ATOCore
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
Files
7650c339a2c879d87f0ad90e0dc428a79a71d387
ATOCore/scripts/eval_data/candidate_queue_2026-04-12.json

1 line
9.0 KiB
JSON
Raw Normal View History

feat: Day 3 — auto-triage via LLM second pass scripts/auto_triage.py: fetches candidate memories, asks a triage model (claude -p, default sonnet) to classify each as promote / reject / needs_human, and executes the verdict via the API. Trust model: - Auto-promote: model says promote AND confidence >= 0.8 AND dedup-checked against existing active memories for the project - Auto-reject: model says reject - needs_human: everything else stays in queue for manual review The triage model receives both the candidate content AND a summary of existing active memories for the same project, so it can detect duplicates and near-duplicates. The system prompt explicitly lists the rejection categories learned from the first two manual triage passes (stale snapshots, impl details, planned-not-implemented, process rules that belong in ledger not memory). deploy/dalidou/batch-extract.sh now runs extraction (Step A) then auto-triage (Step B) in sequence. The nightly cron at 03:00 UTC will run the full pipeline: backup → cleanup → rsync → extract → triage. Only needs_human candidates reach the human. Supports --dry-run for preview without executing. Supports --model override for multi-model triage (e.g. opus for higher-quality review, or a future Gemini/Ollama backend). Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-12 12:30:57 -04:00
{"memories":[{"id":"7d4bdab5-3a8b-4e85-a1b1-117d6d5c094e","memory_type":"project","content":"AtoCore extraction must stay off the hot capture path; batch endpoint only","project":"atocore","confidence":0.6,"status":"candidate","reference_count":0,"last_referenced_at":"","updated_at":"2026-04-12 15:47:16"},{"id":"af4dbc85-e5bd-4aac-927b-5ecee4497ec4","memory_type":"project","content":"Auto-promote gate: confidence ≥0.8 AND no duplicate in active memories","project":"atocore","confidence":0.6,"status":"candidate","reference_count":0,"last_referenced_at":"","updated_at":"2026-04-12 15:47:16"},{"id":"a6d9c14d-e64c-4e9d-aeeb-c9667c5fbda0","memory_type":"project","content":"AtoCore LLM extraction pipeline deployed on Dalidou host, runs via cron at 03:00 UTC via scripts/batch_llm_extract_live.py","project":"atocore","confidence":0.6,"status":"candidate","reference_count":0,"last_referenced_at":"","updated_at":"2026-04-12 15:47:05"},{"id":"06a43be1-026a-4085-b3e1-76348fd78250","memory_type":"project","content":"LLM extractor runs host-side (not in container) because claude CLI not available in container environment","project":"atocore","confidence":0.6,"status":"candidate","reference_count":0,"last_referenced_at":"","updated_at":"2026-04-12 15:47:05"},{"id":"5458852f-6001-44d9-960b-71f6ef5b93b5","memory_type":"project","content":"Host-side extraction script scripts/batch_llm_extract_live.py uses pure stdlib, no atocore imports for deployment simplicity","project":"atocore","confidence":0.6,"status":"candidate","reference_count":0,"last_referenced_at":"","updated_at":"2026-04-12 15:47:05"},{"id":"6ba13dc3-2157-46ec-bd75-9bf9bd37aef1","memory_type":"project","content":"POST /admin/extract-batch accepts mode: rule|llm, POST /interactions/{id}/extract now mode-aware","project":"atocore","confidence":0.6,"status":"candidate","reference_count":0,"last_referenced_at":"","updated_at":"2026-04-12 15:47:05"},{"id":"40db1e3f-c748-40e1-a224-01ad65b0e4fd","memory_type":"knowledge","content":"claude CLI 2.0.60 removed --no-session-persistence flag, extraction sessions now persist in claude history","project":"atocore","confidence":0.6,"status":"candidate","reference_count":0,"last_referenced_at":"","updated_at":"2026-04-12 15:47:05"},{"id":"e66efb99-9b3b-4056-8f65-2dcaa773eb7f","memory_type":"adaptation","content":"Durable memory extraction candidates must be <200 chars, stand-alone, typed as project|knowledge|preference|adaptation","project":"atocore","confidence":0.6,"status":"candidate","reference_count":0,"last_referenced_at":"","updated_at":"2026-04-12 15:47:05"},{"id":"557329d4-ddf1-4b0e-b628-aa29a88a243b","memory_type":"adaptation","content":"Memory extraction confidence defaults to 0.5, raise to 0.6 only for unambiguous committed claims","project":"atocore","confidence":0.6,"status":"candidate","reference_count":0,"last_referenced_at":"","updated_at":"2026-04-12 15:47:05"},{"id":"4ca2545d-9437-48a8-b197-6da6b716f12a","memory_type":"project","content":"Live Dalidou is on commit 39d73e9, not e2895b5","project":"atocore","confidence":0.6,"status":"candidate","reference_count":0,"last_referenced_at":"","updated_at":"2026-04-12 14:59:46"},{"id":"3c3c6cff-7a2b-41d6-b69a-ba12683d4bc2","memory_type":"project","content":"Live harness is reproducible at 16/18 PASS","project":"atocore","confidence":0.6,"status":"candidate","reference_count":0,"last_referenced_at":"","updated_at":"2026-04-12 14:59:46"},{"id":"119a2d0f-2405-45cb-b54c-984ffc12c6ea","memory_type":"project","content":"Live active memories count is 36","project":"atocore","confidence":0.6,"status":"candidate","reference_count":0,"last_referenced_at":"","updated_at":"2026-04-12 14:59:46"},{"id":"4e7e0c38-1080-4631-a8b7-1a1216ed141e","memory_type":"project","content":"Wave 2 project-state entries on live: p04=5, p05=6, p06=6","project":"atocore","confidence":0.6,"status":"candidate","reference_count":0,"last_referenced_at":"","updated_at":"2026-04-12 14:59:46"},{"id":"8304fce1-7bbc-4e55-8c12-c5f127067235","memory_type":"project","content":"R6 is fixed by commit 39d73e9","project":"at
Reference in New Issue Copy Permalink