CookiesForAI

finishing-a-development-branch

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.

Agent skill by Jesse Vincent · Updated

Install Finishing a Development Branch

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 finishing-a-development-branch

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

What Finishing a Development Branch does

Finishing a Development Branch is one of the skills in Superpowers, Jesse Vincent's open-source method for AI coding agents. It loads when implementation is complete and the work needs to be integrated. The agent first runs the project's full test suite. If anything fails, it reports the failures and stops: the options appear only after a green run. Before any merge it also confirms which branch the work split from.

Finishing a Development Branch then checks whether the agent is in a normal repository or a worktree, and shows you a fixed menu: merge back locally, push the branch and open a pull request, or keep the branch as it is. From a detached HEAD there are two options, with no local merge. The integration decision is yours, and the agent waits for it. A local merge is tested again on the merged result before the branch is deleted.

Finishing a Development Branch removes a worktree only if Superpowers created it (under .worktrees/ or worktrees/) and leaves any other worktree to the tool that made it. If removal is refused because of files that were never committed, it lists them and asks you what to do instead of forcing it. Discarding the work happens only when you ask, and only after you type the word discard. It force-pushes only on your explicit request.

When to use Finishing a Development Branch

  • Your agent finished a feature and you want to choose between merging it and opening a pull request.
  • You want the tests run on the final tree before anything is merged.
  • You use git worktrees and want them cleaned up without losing uncommitted files.

When to pick something else

  • You want a code review before deciding: use Requesting Code Review.

Which agents Finishing a Development Branch works in

Jesse Vincent documents Finishing a Development Branch 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.

Finishing a Development Branch license

Finishing a Development Branch is published under MIT.

  • Code review: code-review is Sentry's agent skill for reviewing pull requests by Sentry's engineering practices: runtime errors, performance, side effects, compatibility, security, design, tests and the tone of feedback. (1,031 repository stars on October 2, 2026)
  • GitHub address comments: gh-address-comments is OpenAI's agent skill for pull request review feedback: the agent fetches every comment and review thread on the open PR, numbers them, and fixes only the ones you pick. (27,839 repository stars on October 2, 2026)
  • GitHub fix CI: gh-fix-ci is OpenAI's agent skill for failing pull request checks: the agent uses the GitHub CLI to find failing GitHub Actions runs, pulls the logs, summarizes the failure and proposes a fix plan. (27,839 repository stars on October 2, 2026)
  • Iterate on PR: iterate-pr is Sentry's agent skill that keeps working on a GitHub pull request until CI passes and high and medium priority review comments are fixed, using bundled scripts and the gh CLI. (1,031 repository stars on October 2, 2026)

More from Jesse Vincent

  • 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)
  • 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)

Lists that include finishing-a-development-branch