Line Cook
Structured AI workflow execution: prep → cook → serve → tidy.
When to Use
- Starting a work session with
/line-run - Running individual workflow steps:
/line-prep,/line-cook,/line-serve,/line-tidy - Managing beads issues during execution
- Understanding workflow guardrails
Quick Reference
| Command | Purpose |
|---|---|
/line-run |
Full prep→cook→serve→tidy cycle |
/line-prep |
Sync git, show ready tasks |
/line-cook |
Claim and execute a task |
/line-serve |
AI peer review of completed work |
/line-tidy |
Commit, sync beads, push |
Core Workflow
/line-run # Full cycle with auto-selected task
/line-run <task-id> # Full cycle with specific task
Step-by-Step
- Prep: Sync state, identify available work
- Cook: Claim task, execute with guardrails, verify completion
- Serve: AI reviews changes for quality
- Tidy: File discovered issues, commit, push
Guardrails
Line Cook enforces these disciplines:
- Sync before work - Always start with current state
- One task at a time - Focus prevents scope creep
- Verify before done - Tests pass, code compiles
- File, don't block - Discovered issues become new beads
- Push before stop - Work isn't done until it's pushed
Beads Integration
Line Cook orchestrates beads for task management:
bd ready # Find unblocked tasks
bd update <id> --status=in_progress # Claim task
bd close <id> # Complete task
bd sync # Sync with git
Parking Lot Pattern
Tasks under "Retrospective" or "Backlog" epics are excluded from auto-selection. Explicit selection still works:
/line-cook <parked-task-id>
Error Handling
If a step fails:
- Prep fails: Fix sync issues, retry
- Cook fails: Continue to tidy to save progress
- Serve fails: Review skipped, continue to tidy
- Tidy fails: Create bead for follow-up
Reference
README.mdin project root - Philosophy and installationcommands/directory - Detailed command documentation
Converted and distributed by TomeVault — claim your Tome and manage your conversions.