Turns your prompt into a single, actionable plan without touching any files. It scans your README, docs, and relevant source to understand your stack and constraints, then asks at most one or two clarifying questions before delivering the plan in its final message. The read-only constraint is smart since it forces you to review before execution, though you'll still need another tool to actually implement. Most useful when you're starting a feature or refactor and want to think through the approach before writing code. The minimal question policy keeps things moving instead of getting stuck in planning paralysis.
npx -y skills add openai/skills --skill create-plan --agent claude-codeInstalls into .claude/skills of the current project.
Select a file.
sickn33/antigravity-awesome-skills
github/awesome-copilot
github/awesome-copilot