| name | despaghettify-orchestrate |
| description | Routes agents to the correct Despaghettify task markdown — check vs solve vs add-task vs autonomous loop vs clean vs reset. Triggers on despaghettify orchestration, spaghetti pipeline, M7 scan then implement, run spaghetti-autonomous-agent-task, which DS next, wipe artefacts only or full hub reset. |
Despaghettify orchestrate (router)
Do not duplicate policy or checklists here. Pick one track and follow only that task file end-to-end:
| Intent | Open and follow (single source) |
|---|
Retune C1..C7 bars, M7 weights, or M7_ref | despaghettify/spaghetti-setup.md |
| Structure scan / M7 / hotspots / DS table when triggers fire | despaghettify/spaghetti-check-task.md |
Implement one DS-* (with run spaghetti-solve-task DS-nnn) | despaghettify/spaghetti-solve-task.md |
| Autonomous loop: solve open input-list DS-* backlog first, then check → solve → check until setup conditions clear or hard stop | despaghettify/spaghetti-autonomous-agent-task.md |
| Plan new DS-* rows for one C1–C7 category | despaghettify/spaghetti-add-task-to-meet-trigger.md |
Wipe only workstream pre/post artefacts (+ optional ephemeral cleanup) | despaghettify/spaghetti-clean-task.md |
| Full hub reset (clean + EMPTY template + one check pass) | despaghettify/spaghetti-reset-task.md |
Trigger values (C1..C7 bars, weights, M7_ref) live only in spaghetti-setup.md — never copy them into this skill.
Optional CLI helpers (metrics / listing / preflight): references/CLI.md — helpers only, not a second policy.
Input list: despaghettify/despaghettification_implementation_input.md (open DS rows only). Closed wave history: despaghettify/despaghettification_completed_log.md. Hub CLI auto-runs archive sync on every subcommand (sync-archive / DESPAG_SKIP_ARCHIVE_SYNC=1 to disable).
