projectBoard-plugin

Use FIRST, when pointed at a repo this team hasn't worked before (or to re-check a repo's readiness) — BEFORE decomposing or dispatching any feature. Scans the repo for the preconditions a coding-agent loop needs, AUTO-FIXES the safe/deterministic gaps (beads workspace, ignored agent scratch, the build/test gate) by delegating to the coder, and BOARDS the judgment gaps (a project grounding doc, PR CI) as features. Ends with a readiness report + a human gate. Do NOT use to plan features or write code — that's `decompose-project` and the coder.

Use to run a retro on the coding loop's own track record and distill the lessons into durable grounding. Mines the board's attempt/outcome history (board_retro) for RECURRING failure classes + flow stats, then (a) WRITES each new lesson to the knowledge graph (domain "loop-lessons") — which the coding loop injects into coder prompts at build time — and (b) PROPOSES the same gotchas for PROTO.md in a dated retro report for a human to apply. The self-improving half of the flywheel. Runs unsupervised on a schedule (the dream/distill pattern) — it PROPOSES, it does not auto-edit PROTO.md or commit; a human applies the report. Does NOT write code.

Use when asked to plan, decompose, or spec out a new project/idea before any code is written. Runs the adversarial decomposition pipeline (idea → outline → MADR decision records → epics › milestones › features with EARS acceptance criteria), writing the plan as docs in the working tree AND populating the beads board — with an antagonist hardening every step and a human approval gate per epic. The board's loop then builds the `ready` features. Do NOT use to write code; this only plans.