This is the planning half of a two-skill workflow that separates design from execution. It walks through context loading, clarification questions, and outputs a timestamped folder under docs/.plans with a SUMMARY.md and numbered phase files. Each phase gets complexity ratings, task checklists, and verification commands. The goal is a plan so self-contained that execute-plan can pick it up in a fresh session, even after /clear. It's opinionated about granularity (2-10 minute tasks), includes research folders when needed, and enforces a "no coding during planning" rule. Best for refactors, migrations, or anything risky enough that you want a reviewer or your future self to understand the strategy before touching production files.
npx -y skills add buiducnhat/agent-skills --skill write-plan --agent claude-codeInstalls into .claude/skills of the current project.
Select a file.
sickn33/antigravity-awesome-skills
moizibnyousaf/ai-agent-skills
github/awesome-copilot