brewcode:start

Execute Task — [task-file-path]

How It Works

/brewcode:start → Load PLAN.md → Execute phases
   ↓
PreToolUse(Task)  → Inject ## K knowledge + protocol reminder
   ↓
PostToolUse(Task) → Remind: WRITE report → CALL coordinator
   ↓
PreCompact        → Validate + compact KNOWLEDGE → handoff
   ↓
Re-read PLAN.md   → Continue from current phase

Execution Steps

1. Resolve Task Path

• If $ARGUMENTS has path → use it
• If $ARGUMENTS empty → read .claude/TASK.md (first line = active path)
• If neither → STOP: ❌ No task path! Run: /brewcode:spec "description" then /brewcode:plan

2. Initialize via Coordinator (REQUIRED)

Task tool:
  subagent_type: "brewcode:bc-coordinator"
  prompt: "Mode: initialize. Task path: {TASK_PATH}"

Coordinator validates, creates lock, updates status → in progress.

3. Load Context

• Read task file content
• Read KNOWLEDGE.jsonl if exists
• Verify reports directory exists

4. Execute Phases

FOR each phase:
  1. Read phase requirements
  2. Call agent (developer/tester/reviewer)

  ⛔ MANDATORY (after EACH agent):
  3. WRITE report → artifacts/{P}-{N}{T}/{AGENT}_output.md
  4. CALL bc-coordinator → reads report, extracts knowledge

  5. Run verification phase (same 2-step protocol)
  6. Iterate or proceed based on results

5. Complete

• Status → finished
• Extract rules (REQUIRED): Skill(skill="brewcode:rules", args="{KNOWLEDGE_PATH}")