Use when you have a written implementation plan to execute in a separate session with review checkpoints
npx skills add terry-li-hm/skills --skill "plan"
Install specific skill from multi-skill repository
# Description
Transform feature descriptions into well-structured project plans following conventions. Research, validate, and create actionable plans.
# SKILL.md
name: plan
description: Transform feature descriptions into well-structured project plans following conventions. Research, validate, and create actionable plans.
Plan
Transform feature descriptions, bug reports, or improvement ideas into well-structured markdown plans.
Trigger
Use when:
- User says "plan this" or "create a plan for"
- Starting a new feature
- User has a feature description or idea
Inputs
- Feature description (required, or ask if missing)
- Plan detail level (minimal/standard/comprehensive - auto-detect based on complexity)
Workflow
Phase 1: Idea Refinement
Check for brainstorm documents first:
ls -la docs/brainstorms/*.md 2>/dev/null | head -10
If a relevant brainstorm exists (matching topic, within 14 days):
1. Read it
2. Announce: "Found brainstorm from [date]. Using as context."
3. Skip refinement questions
If no brainstorm, refine via dialogue:
Ask questions one at a time:
- Purpose: What problem does this solve?
- Constraints: Any hard requirements?
- Success criteria: How do we know it's done?
Continue until clear or user says "proceed".
Phase 2: Research
Run parallel subagents:
Task explore("Find relevant files, patterns, and conventions in this codebase for: {feature_description}")
Task explore("Search docs/solutions/ and docs/learn/ for related learnings to: {feature_description}")
Decide on external research:
- High-risk (security, payments, external APIs) → always research
- Strong local context → skip external
- Uncertainty → research
If external research needed:
Task general("Research best practices for: {feature_description}. Search for current patterns and gotchas.")
Phase 3: Plan Structure
Choose detail level:
| Level | Best For | Sections |
|---|---|---|
| minimal | Simple bugs, small improvements | Problem, Acceptance Criteria, Context |
| standard | Most features, complex bugs | Overview, Problem, Solution, Technical Considerations, Acceptance Criteria |
| comprehensive | Major features, architecture changes | Full structure with phases, alternatives, risk analysis |
Auto-detect: If feature description > 100 words or mentions multiple systems → comprehensive. Otherwise standard.
Phase 4: Create Plan File
Filename: docs/plans/YYYY-MM-DD-<type>-<descriptive-name>-plan.md
Example: 2026-01-29-feat-user-auth-flow-plan.md
Template (standard level):
---
title: [Issue Title]
type: [feat|fix|refactor]
date: YYYY-MM-DD
---
# [Issue Title]
## Overview
[Brief description]
## Problem Statement / Motivation
[Why this matters]
## Proposed Solution
[High-level approach]
## Technical Considerations
- Architecture impacts
- Performance implications
- Security considerations
## Acceptance Criteria
- [ ] Requirement 1
- [ ] Requirement 2
- [ ] Testing requirements
## References
- Similar code: [file_path:line_number]
- External docs: [url]
Output
- Write plan to
docs/plans/YYYY-MM-DD-*-plan.md - Present options:
- Open in editor
- Run
/workto start implementing - Create GitHub issue
Examples
User: "Plan a user authentication system"
→ Creates docs/plans/2026-01-29-feat-user-auth-plan.md
User: "Plan: fix the race condition in checkout"
→ Creates docs/plans/2026-01-29-fix-checkout-race-plan.md
# Supported AI Coding Agents
This skill is compatible with the SKILL.md standard and works with all major AI coding agents:
Learn more about the SKILL.md standard and how to use these skills with your preferred AI coding agent.