A discipline layer for Claude Code. Hooks at every tool boundary, a workflow that runs from intake to commit, and a small constitution the agent cannot bypass.
Command Claude Code
Record human approval of a workflow's direction (the intake + its acceptance criteria + the CO-A evidence). The Direction Approval Guard hook blocks Claude from ever writing approval tokens; this command is the only sanctioned path. Must be user-invoked. Replaces the retired /approve-spec gate (D3/CO-E gate-collapse)…
Command Claude Code
Record human approval of a swarm plan. swarm-dispatch will not run until this approval token exists. Must be user-invoked — Claude cannot approve.
Command Claude Code
Grant consent for Claude to run git commit. Workflow-scoped when a workflow is active, else a time window. Required by the Git Commit Guard hook.
Command Claude Code
Grant consent for Claude to run git push. Valid for 5 minutes. Required by the Git Commit Guard hook on protected branches.
Command Claude Code
Detect and repair baseline drift — missing or invalid .claude/workflows.jsonl, schema violations, four-way Article IV / §18 mirror drift, and (advisory) shipped-tooling files placed outside .claude/ against the convention codified at seed.md §3. Interactive: presents each detected violation via AskUserQuestion and…
Command Claude Code
Configure the baseline for this specific project. Invokes the scout skill, then claude-automation-recommender, populates .claude/project.json, pre-creates lazy directories, re-renders swarm-worker.md from its template with any stack-specific skills appended, appends a §16 addendum to docs/init/seed.md, runs…