Plan Iteration

Steps

1. Read planning/active/new-iteration.md (error if missing: "Run /pick-from-backlog or /add-feature first."). Parse all items.

2. Gather context: Read docs/overview.md, planning/plan-overview.md (if exists), and use Glob to explore project structure.

3. Check for existing plan: If planning/active/plan-iteration.md exists, ask: "Regenerate or update?"

4. Determine iteration number from new-iteration.md title. If not found:

   python "${CLAUDE_PLUGIN_ROOT}/scripts/get-next-iteration.py"
   

5. Generate detailed plan. For each item: break into steps, identify files to create/modify, note dependencies, estimate complexity (Low/Medium/High), consider risks.

6. Create planning/active/plan-iteration.md:

   # Iteration XXX Implementation Plan
   
   ## Iteration Goals
   - Goal 1 (from #XXX)
   - Goal 2 (from #YYY)
   
   ## Implementation Steps
   
   ### Step 1: [Task Name]
   **Related to:** #XXX
   **Files affected:**
   - `path/to/file1.ts` (create)
   - `path/to/file2.ts` (modify)
   **Description:** [What to do and how]
   **Dependencies:** None / Step X
   **Complexity:** Low | Medium | High
   **Risks:** [If any]
   
   ### Step 2: [Task Name]
   ...
   
   ## Testing Strategy
   - [Test approaches]
   
   ## Documentation Updates
   - [Files needing doc updates]
   
   ## Overall Risks
   - [Risks and mitigations]
   

7. Run consistency check:

   python "${CLAUDE_PLUGIN_ROOT}/scripts/consistency-check.py"
   

8. Output: Confirm plan, show step count and main goals. Recommend:

   git add planning/ && git commit -m "Plan iteration XXX"
   

   Next: /implement-iteration

Notes

• Be SPECIFIC: include actual file paths, not placeholders.
• A good plan should be executable by someone unfamiliar with prior decisions.
• Order steps by dependencies. Break complex tasks into smaller steps.