/getting-started

Output this guide to the user. Do not act on it - display it for reference.

---

How Line Cook Works

Line Cook organizes AI-assisted development into two cycles: Mise (planning) and Run (execution). Think big in Mise, execute small in Run.

The Two Cycles

  Mise Cycle                  Run Cycle
 ┌──────────────────┐       ┌──────────────────┐
 │                  │       │                  │
 │   /brainstorm    │       │   /prep ◄────┐   │
 │        ↓         │       │      ↓       │   │
 │     /scope       │       │   /cook      │   │
 │        ↓         │       │      ↓       │   │
 │   /finalize ─────┼─────► │   /serve     │   │
 │                  │       │      ↓       │   │
 └──────────────────┘       │   /tidy ─────┘   │
                            │    next task     │
                            └──────────────────┘

Mise Cycle: Ideas → Tasks

Turn unstructured ideas into well-scoped, dependency-ordered tasks.

Phase	Command	What it does
Brainstorm	/line:brainstorm	Explore the problem space — questions, research, risks
Scope	/line:scope	Structure into epics → features → tasks with dependencies
Finalize	/line:finalize	Create beads (tracked issues) and test specifications

Use /line:mise to run all three with pauses between phases.

Run Cycle: Tasks → Shipped Code

Execute one task at a time with TDD, AI review, and automatic commit/push.

Phase	Command	What it does
Prep	/line:prep	Sync git/beads, show unblocked tasks
Cook	/line:cook	Claim task, write tests first, implement, close
Serve	/line:serve	AI peer review (polisher + sous-chef)
Tidy	/line:tidy	File discoveries, commit, push

Use /line:run to run all four in sequence.

Quality Gates

After the basic cycle, additional gates validate completed work:

Gate	Command	When
Feature done	/line:plate	All tasks under a feature are closed
Epic done	/line:close-service	All features under an epic are closed

Workflow Principles

1. Sync before starting — Always work from current state
2. Track with beads — Strategic tasks live in the issue tracker
3. Note, then file — Discoveries are noted during cook, filed during tidy
4. Guardrails on completion — Verify before marking done
5. Push before stop — Session isn't done until pushed

Tip: Run /line:plan-audit periodically to check bead health and structure.

What's Next?

Ready to try it hands-on? Run /line:onboarding for an interactive walkthrough with a demo project.

Having setup trouble? Run /line:init to verify your environment.

Already set up?

• Start planning: /line:mise
• Start executing: /line:run
• See all commands: /line:help
• Autonomous execution: /line:loop