CookiesForAI

writing-plans

Writing Plans

Writing Plans is an agent skill from Superpowers: your agent turns a spec into a step-by-step implementation plan with exact files, signatures and tests, before touching code.

Agent skill by Jesse Vincent · Updated

Install Writing Plans

In Claude Code, install the whole Superpowers plugin from Anthropic's official marketplace

Terminal
/plugin install superpowers@claude-plugins-official

Or install only this skill with the skills CLI

Terminal
npx skills add obra/superpowers --skill writing-plans

Or paste this into your coding agent: Install the agent skill writing-plans from github.com/obra/superpowers. A skill can include scripts that run on your computer, so read its source first.

What Writing Plans does

Writing Plans is one of the skills in Superpowers, Jesse Vincent's open-source method for AI coding agents. It loads once a multi-step task has a spec or a set of requirements and no code has been written yet. The agent writes the plan for an engineer who has never seen the codebase or the spec, recording the decisions that engineer could not make alone: the files to touch, the names and signatures to use, the values taken from the spec, and the tests that prove each task.

Writing Plans saves plans to docs/superpowers/plans/ unless you prefer another place. Each starts with a header naming the goal, architecture, tech stack and spec, a Global Constraints section with values copied from the spec, and a Review Focus list of failure modes no test covers yet. Every task lists its exact files and the interfaces it consumes and produces, split into small checkbox steps: write the failing test, run it, implement, run it again, commit.

Before Writing Plans hands over, the agent checks the plan against the spec for requirements with no task, steps that decide nothing, names that differ between tasks, and a plan far longer than its spec. It then asks you to review the plan and choose how to run it: Subagent-Driven Development, where a fresh subagent does each task, or native execution with Executing Plans. An isolated worktree, if used, comes from Using Git Worktrees.

When to use Writing Plans

  • You have a spec or design and want it broken into small, testable tasks before any code is written.
  • You plan to hand the work to subagents or another session and need a plan that stands on its own.
  • Your agent's plans say things like "handle edge cases" and you want exact files and tests instead.

When to pick something else

  • You do not have a spec or design yet: start with Brainstorming.

Which agents Writing Plans works in

Jesse Vincent documents Writing Plans for Claude Code, Codex (app and CLI), Cursor, Gemini CLI, GitHub Copilot CLI, OpenCode, Antigravity, Devin CLI, Factory Droid, Grok Build CLI, Kimi Code, Pi, Qwen Code, Hermes Agent, Muse. The open skills CLI also installs it into 78 agents, including Claude Code, Codex, Cursor, Gemini CLI, GitHub Copilot, OpenCode (we listed it with the CLI on October 1, 2026). See where each agent looks for skills.

Superpowers is written in Claude Code's vocabulary (subagent dispatch, todos, the Skill tool) and ships tool mappings for Codex, Gemini CLI, Pi, Antigravity, Hermes Agent and Muse. Installed as the full plugin, a session-start hook loads Using Superpowers so the agent reaches for skills on its own; one skill installed with the skills CLI gets no hook and may point to sibling skills you do not have.

Writing Plans license

Writing Plans is published under MIT.

  • Grill me: grill-me is Matt Pocock's agent skill that turns your agent into a relentless interviewer, asking numbered rounds of questions about a plan or design until every decision is settled. (273,901 repository stars on October 2, 2026)
  • Brainstorming: Brainstorming is an agent skill from Superpowers that makes your coding agent work out what you want and agree a design with you before it writes any code. (293,976 repository stars on October 2, 2026)
  • Diagnosing Superpowers: Diagnosing Superpowers is an agent skill from Superpowers that reads a session's transcripts when work went wrong and reports what happened, with every finding tied to a file and line. (293,976 repository stars on October 2, 2026)
  • Dispatching Parallel Agents: Dispatching Parallel Agents is an agent skill from Superpowers: when several problems are independent, your agent sends a focused subagent to each at once, then checks the results together. (293,976 repository stars on October 2, 2026)

More from Jesse Vincent

  • Executing Plans: Executing Plans is an agent skill from Superpowers in which your agent works through a written plan itself, task by task with tests, then gets one fresh review of the whole branch. (293,976 repository stars on October 2, 2026)
  • Finishing a Development Branch: Finishing a Development Branch is an agent skill from Superpowers: when work is done, your agent runs the tests, offers to merge, open a pull request or keep the branch, then cleans up. (293,976 repository stars on October 2, 2026)
  • Receiving Code Review: Receiving Code Review is an agent skill from Superpowers: your agent checks review feedback against the code before acting, pushes back with reasons when it is wrong, and skips the flattery. (293,976 repository stars on October 2, 2026)
  • Requesting Code Review: Requesting Code Review is an agent skill from Superpowers: your agent sends its changes to a separate reviewer subagent that checks them against the plan and grades each issue by severity. (293,976 repository stars on October 2, 2026)

Lists that include writing-plans