janpeter/Scrum4Me
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

docs: update ClaudeJob lifecycle with plan_snapshot

Browse source 
Document state machine snapshot capture/reset, plan_snapshot field
rationale, and drift-detection baseline semantics.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
This commit is contained in:
Janpeter Visser 2026-04-30 19:27:37 +02:00
parent a3af2dda63
commit 075af9d619
1 changed files with 7 additions and 3 deletions
Download patch file Download diff file Expand all files Collapse all files

10
docs/scrum4me-architecture.md
View file

@ -1120,12 +1120,15 @@ Developers kunnen vanuit de Task Detail Dialog een lokale Claude Code-sessie ins
### State machine
```
QUEUED → CLAIMED → RUNNING → DONE
→ FAILED
QUEUED → CLAIMED (snapshot capture) → RUNNING → DONE
→ FAILED
→ CANCELLED (door user)
CLAIMED → QUEUED (stale claim cleanup, >30min)
CLAIMED → QUEUED (stale claim cleanup, >30min; snapshot gewist)
QUEUED → CLAIMED (re-claim na stale reset; snapshot refreshed)
```
**Snapshot-rationale:** bij atomic claim schrijft `wait_for_job` de dan-actuele `task.implementation_plan` naar `claude_jobs.plan_snapshot`. Dit veld blijft bevroren terwijl de job loopt — ook als een gebruiker `update_task_plan` aanroept. Zo kan een toekomstige verify-tool drift detecteren tussen de baseline (snapshot) en de actuele plan. Jobs zonder snapshot (NULL) zijn aangemaakt vóór deze feature en worden als "no baseline" gemarkeerd.
### ClaudeJob model
```
@ -1134,6 +1137,7 @@ claude_jobs
status: ClaudeJobStatus (QUEUED|CLAIMED|RUNNING|DONE|FAILED|CANCELLED)
claimed_by_token_id (FK → api_tokens, nullable)
claimed_at, started_at, finished_at
plan_snapshot: String? — bevroren snapshot van task.implementation_plan bij claim
branch, summary, error
@@index([user_id, status])
@@index([task_id, status])