crank
Fully autonomous epic execution. Runs until ALL children are CLOSED. Loops through beads issues, runs /implement on each, validates with /vibe. NO human prompts, NO stopping.
npx skills add majiayu000/claude-skill-registry --skill crank --agent claude-code
Same command for any agent — swap --agent for codex, cursor, copilot.
Weekly change comes from our own snapshots, not the repository page — it measures attention, not adoption.
# Crank Skill > **Quick Ref:** Autonomous epic execution. Loops `/implement` on all issues until DONE. Output: closed issues + final vibe. **YOU MUST EXECUTE THIS WORKFLOW. Do not just describe it.** Autonomous execution: implement all issues until the epic is DONE. **Requires:** bd CLI (beads) for issue tracking, OR in-session TaskList for task-based tracking. ## Global Limits **MAX_EPIC_ITERATIONS = 50** (hard limit across entire epic) This prevents infinite loops on circular dependencies or cascading failures. **Why 50?** - Typical epic: 5-10 issues - With retries: ~5 iterations per issue max - 50 = safe upper bound (10 issues × 5 retries) **MAX_PARALLEL_AGENTS = 3** (hard limit per wave) When multiple issues are ready, execute them in parallel using subagents. Capped at 3 to prevent context explosion. **Why 3?** - Each subagent returns results that accumulate in context - 3 parallel agents = manageable context growth - Higher parallelism risks context overflow on complex issues ## Completion Enforcement (The Sisyphus Rule) **THE SISYPHUS RULE:** Not done until explicitly DONE. After each task, output completion marker: - `<promise>DONE</promise>` - Epic truly complete, all issu
- Global Limits
- Completion Enforcement (The Sisyphus Rule)
- Execution Steps
- Step 0: Load Knowledge Context (ao Integration)
- Step 1: Identify the Epic
- Step 1a: Initialize Iteration Counter
- Step 2: Get Epic Details
- Step 3: List Ready Issues (Current Wave)
- Step 3a: Pre-flight Check - Issues Exist
- Step 4: Execute Wave (Parallel Subagents)
- Step 5: Track Progress (No Per-Issue Vibe)
- Step 6: Check for More Work
- Step 7: Final Batched Validation
- Step 8: Extract Learnings (ao Integration)
If ao CLI available, inject prior knowledge about epic execution if command -v ao &>/dev/null; then ao search "epic execution implementation patterns" 2>/dev/null | head -20 ao flywheel status 2>/dev/null ao ratchet status 2>/dev/null fi bd list --type epic --status open 2>/dev/null | head -5 ls -lt .agents/plans/ 2>/dev/null | head -3 Initialize crank tracking in epic notes bd update <epic-id> --append-notes "CRANK_START: iteration=0 at $(date -Iseconds)" 2>/dev/null
What does the crank skill do?
Fully autonomous epic execution. Runs until ALL children are CLOSED. Loops through beads issues, runs /implement on each, validates with /vibe. NO human prompts, NO stopping.
How do I install it?
Run `npx skills add majiayu000/claude-skill-registry --skill crank --agent claude-code` — it drops the skill into your project so the agent can pick it up. Swap the --agent value for codex, cursor or copilot if you use one of those.
Where does this skill come from?
From majiayu000/claude-skill-registry, a repository with 534 stars. We read it straight from the repository tree rather than a submitted listing, so what you see here is what is actually published.
Is a popular skill a good skill?
Not necessarily. Stars measure attention, not adoption — a repository can trend for a week and be abandoned. That is why we show the weekly change from our own snapshots next to the total, instead of a single flattering number.