executing-plans Skill for Claude: What It Does and How to Install It

Makes Claude carry out a written implementation plan task by task in one session, with a test gate per task and one review at the end.

Installs: 250.3K; on skills.sh
Stars: 310.6K; on GitHub
Files: 2; besides SKILL.md
Updated: Oct 2026; last repo commit

executing-plans makes Claude run a plan itself, inline, instead of dispatching a subagent for every task. It keeps a ledger file so it can resume after its context is compacted, uses test-driven development as the per-task gate, and has a fresh reviewer check the whole branch at the end (or reviews it itself, and says so, if there is no subagent tool). It does not pause between tasks to ask whether to continue. It records its own decisions as rulings in the ledger.

Only four things stop it: irreversible or destructive actions, security-sensitive actions, side effects outside the worktree such as a merge or a push to a shared branch, and a plan so broken that every path is a guess. It is built for plans written with the writing-plans skill, and it calls other superpowers skills and shell scripts from the same plugin, so install the whole superpowers set.

Good for

Developers who already have a written plan
People who want lower cost than a subagent per task
Setups where no subagent tool is available

Install executing-plans

Claude Code

Copy this prompt.

Install the executing-plans skill in Claude Code. Run this command:

/plugin install superpowers@claude-plugins-official

Then check that it loaded and tell me how to use it.

2
Paste it into Claude Code. It shows the change and asks before running it.

This installs the whole superpowers plugin. executing-plans calls sibling skills and scripts from the same plugin, so installing it alone leaves gaps.

From the project README. Source.

What's in the skill

When it loads

Use when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available

Sections in SKILL.md

When to Use
The Process
Setup
The Task Loop
1. Take the task
2. Work the steps
3. The completion contract
4. Complete the task
Final Review
Finish
Common Rationalizations

Key facts

Author: obra
Repo: obra/superpowers (skills/executing-plans)
Installs on skills.sh: 250,310 (Nov 2026)
GitHub stars: 310,635 (Nov 2026)
License: MIT
Last repo commit: Oct 29, 2026
Files in the skill folder: 3 including SKILL.md
skills.sh audits: Gen Agent Trust Hub: pass, Socket: pass, Snyk: pass
Needs: Other superpowers skills and the sdd-workspace and review-package scripts from subagent-driven-development
Own scripts: scripts/task-start and scripts/task-done

What it includes

Works in an isolated worktree and creates a plan workspace with a progress ledger
Runs code · Setup
Script that prints the task brief path and the base commit
Runs code · task-start
Follows the plan with TDD, runs each verification and reads the output
Runs code · Work the steps
Script that runs your test command for the task and logs the result in the ledger
Runs code · task-done
A fresh reviewer checks the whole branch; self-review if no subagent tool
Runs code · Final review
Hands off to the finishing-a-development-branch skill
Finish

From the official docs

Permissions

Medium risk

Claude runs bundled scripts, your tests and git commits across the whole plan without checking in between tasks. Run it on a branch in a worktree and review the commits afterwards.

What it can do

Read your data
Create and change things
Delete things
Read-only mode

Watch out for

It keeps going without checking in between tasks
It stops only for destructive or irreversible actions, security-sensitive actions, shared side effects and broken plans. Start from a clean branch.
It runs bundled scripts and your test command
The task scripts run your tests and write a ledger folder in the repo. Check that your test command is safe.
It makes commits as the plan says
Commit steps in the plan are carried out. Review the commits before you push.
It deletes the plan's workspace
When the final review is clean and its fixes are committed, SKILL.md tells Claude to delete the plan's workspace directory (the git-ignored ledger and briefs under .superpowers/sdd/).

Questions

How do I install executing-plans?

Use the install box above. It installs the whole superpowers set, because this skill calls sibling skills and scripts such as subagent-driven-development, using-git-worktrees and finishing-a-development-branch.

executing-plans or subagent-driven-development?

Pick executing-plans when you chose inline execution or have no subagent tool. Pick subagent-driven-development for a fresh implementer and a review after every task. The second costs a fresh context per task and per review.

Does it need a plan first?

Yes. SKILL.md says it is for a plan from the writing-plans skill, and it reads the plan and its spec before starting.

Will it stop and ask me questions?

Rarely. It decides conflicts itself and logs them in the ledger. It stops only for destructive or irreversible actions, security-sensitive actions, shared side effects like a push or merge, or a plan that is too broken to follow.

Is it safe?

It runs scripts, tests and git commits in your repo and keeps going without checking in. Run it on a branch in a worktree and review the ledger and commits afterwards.

What if the session loses its place?

Progress lives in a ledger file, not only in the chat. SKILL.md tells Claude to trust the ledger and git log after a context compaction.

Information

updated Nov 2, 2026

Repository: obra/superpowers
Folder: skills/executing-plans

obra/superpowers skills.sh

Sources

Docs, issue trackers and discussions this page is built from.

Show all 2

obra/superpowers README · github.com
executing-plans SKILL.md · github.com