mirror of
https://github.com/awizemann/scarf.git
synced 2026-05-10 18:44:45 +00:00
Alan Wizemann
cf8a737ace
Each plan is the product of one Opus 4.7 1M-context planning agent dispatched in parallel to investigate a single v0.13 surface. Plans are read-only — they identify file-by-file edits, type additions, capability gates, tests, and open questions, but no code is written yet. Implementation agents will be dispatched per stream once the cross-stream open-questions matrix is resolved. - WS-2: Persistent Goals + ACP /queue slash commands - WS-3: Kanban v0.13 diagnostics + recovery UX - WS-4: Curator archive + prune + list-archived - WS-5: Messaging Gateway v0.13 expansion (Google Chat, allowlists, behavior toggles) - WS-6: Provider catalog refresh + image_gen.model + OpenRouter response caching - WS-7: Settings tab additions (MCP SSE, Cron --no-agent, Web Tools split, Profiles --no-skills) - WS-8: UX polish (compression count, /new <name>, redaction default, display.language) - WS-9: ScarfGo iOS catch-up (read-only mirrors of WS-2/3/4/5) COORDINATOR-REVIEW.md compiles the cross-stream collision matrix, the open- questions matrix (54 questions clustered into wire-shape unknowns, architectural decisions, and out-of-scope deferrals), and the recommended sequencing for implementation + review. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>