Managing Feature Plans

---

name: managing-feature-plans description: Create, review, update, and archive feature plan documents. Use when starting a new feature, assessing progress, refining requirements, or marking a plan complete. owner: eng-product-process version: 1.1 last_updated: 2026-02-01

When to use this skill

• Start, scope, or refine a new feature plan.
• Review progress, unblock development, or update an existing plan.
• Audit a plan for missing requirements, unclear TODOs, or missing references.
• Mark a plan as complete and archive it.

Do not use this skill for

• Project-level roadmaps or long-lived strategy documents (use the roadmap-skill).
• Ad-hoc notes or meeting minutes that won't be maintained as a plan.

Why use this skill

• Standardizes feature plans in doc/plan/ so they are discoverable and actionable.
• Keeps TODOs, milestones, and acceptance criteria explicit and trackable.
• Automates safe archival of completed plans.

Key features

1. Create a feature plan

• Create doc/plan/{feature-name}-plan.md from doc/plan/plan-template.md (or fallback template).
• Enforce kebab-case names: {feature-name}-plan.md (e.g. storage-interface-plan.md).

2. Review and update

• Detect missing or ambiguous requirements, TODOs, acceptance criteria, and references.
• Edit plan sections: Goals, Non-goals, Requirements, Design, Milestones, Tasks, Risks, Dependencies.
• Maintain TODO/DONE lists and a changelog for completed items (with timestamps where applicable).

3. Archive

• Verify completion (no open TODOs, acceptance criteria satisfied) and move the plan to doc/plan/archived/{feature-name}-plan.md.
• Add a completion stamp (date + short summary) to the archived file.
• Helper: python archive_plan.py {feature-name}

How it works (runtime behavior)

1. Resolve feature name

• Normalize user input to kebab-case and validate uniqueness.
• If ambiguous, propose clearer names and ask for confirmation.

2. Locate or create plan

• If doc/plan/{feature-name}-plan.md exists: open it for review and updates.
• If missing: create it from doc/plan/plan-template.md or a minimal inline template and populate the header.

3. Review and edit

• Ensure sections are present and actionable (Goals, Non-goals, Requirements, Acceptance Criteria).
• Convert vague items into concrete TODOs with owners, estimates, and explicit acceptance criteria where possible.
• Move completed checklist items to the changelog with timestamps and author.

4. Archive when complete

• Preconditions: no open TODOs blocking release, all acceptance criteria marked PASS (or linked to verification evidence), and changelog updated.
• Run: python archive_plan.py {feature-name}
  • This moves the file to doc/plan/archived/{feature-name}-plan.md, appends a completion stamp, and records the archiver.

Acceptance checklist (before archiving)

• [ ] No open TODOs assigned to the feature team.
• [ ] All acceptance criteria marked PASS or have linked verification evidence (tests, design reviews).
• [ ] Milestones and release notes updated.
• [ ] Changelog contains all completed items with timestamps.

Template locations

• Active plans: doc/plan/
• Templates: doc/plan/plan-template.md
• Archived plans: doc/plan/archived/

Usage examples (explicit)

• Create a plan from a prompt: "Create a plan for "Improved upload UX""
  • Resulting filename: doc/plan/improved-upload-ux-plan.md
• Review an existing plan: "Review the plan for storage-interface and list missing requirements"
• Archive when done: "Archive the feature plan upload-progress once tests are green" or run locally: python archive_plan.py upload-progress

Requirements

• Python 3.7+ for archive_plan.py
• Read/write access to doc/plan/ and doc/plan/archived/

Notes

• Keep SKILL.md concise: frontmatter name/description are used to match user intents.
• Prefer concrete suggestions for naming and TODOs to reduce back-and-forth.
• Update the frontmatter owner/version/last_updated when making future changes.