What does this skill do, and when should you use it?
This skill edits the title and body of one or more GitHub pull requests. It can infer a PR from the current branch or commit, or work from an explicitly specified PR, while checking the existing body for information that must be preserved. The rewritten body should explain why the change is being made, what the net change is, and how it was verified. It supports ordinary Git workflows, stacked pull requests, and Sapling-managed repositories.
Reads the current Git branch or commit and can use gh pr view to find its associated pull request; uses sl log or sl sl to inspect the associated GitHub pull request in Sapling repositories; checks the existing PR title and body; uses gh to update the title and body; preserves important existing content such as images; and writes the motivation, net change, relevant references, and verification details.
- A developer needs to add the motivation and verification details to the pull request for the current branch.
- A maintainer needs to revise the titles and bodies of one or more pull requests consistently.
- A team using stacked pull requests needs each body to describe only the net change relative to its actual base branch.
- A Sapling user needs to identify the GitHub pull request associated with the current revision.
How do you install this skill?
- Execution has an external side effect: gh may directly modify a remote PR, but the skill does not explicitly require confirmation, backup, or rollback before editing.
- The core flow depends on git/gh/sl, valid authentication, and network access; failure handling is underspecified when tools are unavailable, a PR cannot be resolved, or permissions are insufficient.
- Chinese-language conventions and reachability or fallback behavior for mainland-China network restrictions are not documented.
- Shell / CLI
- Network access
- Local filesystem
GitGitHub CLI (gh)Sapling SCM (only for Sapling-managed repositories)
The repository README does not document a standalone installation command for this skill. The skill file is located at .codex/skills/codex-pr-body/SKILL.md; place that directory in an Agent Skills-compatible client to make it available. The README documents installation of the overall Codex CLI, but does not say that this installs the skill separately.
tmp="$(mktemp -d)"
git clone --depth 1 https://github.com/openai/codex.git "$tmp"
mkdir -p ~/.claude/skills
cp -R "$tmp/.codex/skills/codex-pr-body" ~/.claude/skills/
rm -rf "$tmp"Generated from the source repository and skill path; it copies only this skill's folder. If the author's install steps above differ, follow those first. To scope it to one project, replace ~/.claude/skills with that project's .claude/skills.
How do you use this skill?
In a client where the skill is available, use a request such as: Update the title and body of the pull request for my current branch, preserving existing images and explaining why, what changed, and how it was verified. For ordinary Git, the skill may combine git branch with gh pr view <branch> --repo openai/codex --json number --jq '.number' to find the PR. If .git/sl/store exists, use the documented Sapling inspection flow.
What are this skill's strengths and limitations?
- Covers PR discovery, existing-body review, and title/body editing in one workflow.
- Explicitly protects existing images and other important information during rewriting.
- Handles ordinary Git, stacked pull requests, and Sapling scenarios.
- Produces collaboration-oriented bodies focused on motivation, net change, and verification.
- Requires the GitHub CLI `gh` and network access; it is not a text-only drafting skill.
- The source does not specify detailed handling for authentication failures, missing pull requests, or permission errors.
- No skill-specific test suite or verification results are provided in the source.
- Its scope is limited to PR titles and bodies; it does not modify code or run tests.
How does this skill compare with similar options?
Side by side with related skills; every score comes from the same FSRS standard.
| Skill | FS score | Stars | Last updated | License |
|---|---|---|---|---|
| Codex Pull Request Editor this page ✓ OpenAI · Official | 50 · Use with care | ★ 129k | 11d ago | Apache-2.0 |
| Yeet GitHub PR Flow ✓ OpenAI · Official | 41 · Not recommended | ★ 28k | 4mo ago | — |
| CODEOWNERS PR Splitter ✓ NVIDIA · Official | 57 · Use with care | ★ 3.6k | 11d ago | Apache-2.0 |
| Ship a PR (e2e repo PR delivery workflow) | 59 · Recommended | ★ 10k | 11d ago | Apache-2.0 |
| RenderCV PR Reviewer | 55 · Use with care | ★ 18k | 7mo ago | MIT |
How did FollowSkills review this skill?
The skill explicitly uses gh to edit PR title and body, requires checking and preserving existing images, and prohibits local absolute paths and confidential information, showing useful data-flow boundaries. It does not require user confirmation, least-privilege configuration, pre-change backup, or rollback, so points are deducted.
The document covers ordinary Git, Sapling, stacked PRs, and body-rewrite rules, including preservation of important content. However, it depends on git, gh, sl, and remote PR state, without diagnostic or recovery guidance for missing tools, absent PR associations, permission failures, or command errors; static calibration keeps this at or below 10.
The target scenario and main inputs—an explicit PR or the current branch/commit—are reasonably clear, and stacked Sapling workflows are addressed. Non-fit cases, exclusion triggers, Windows or Chinese-language support, and behavior in network-restricted environments are not specified, so points are deducted.
The front matter is concise and the sections are readable, with guidance for safe writing, stacked PRs, and Sapling. It lacks installation/dependency notes, a parameter contract, examples, FAQs, versioning, changelog, maintenance ownership, and update path. Apache-2.0 licensing and OpenAI provenance are visible only from repository context, so points are deducted.
The skill directly targets updating PR titles and bodies and specifies why/what ordering, net-change scope, verification notes, Markdown conventions, and documentation follow-up. Its output value is clear, but there is no output template, post-edit verification, or usable failure result, and static review cannot establish successful execution; the score is capped at 7.
The source is auditable and provides textual support for its key behaviors. It contains no skill-specific tests, CI coverage, execution records, or third-party reproduction evidence, so only limited static verifiability is warranted.
Open a dimension to read why it scored that way
Evidence confidence：Low — Mostly static review, author material or a limited demo; useful for discovery, not high-risk decisions.
See the full review method →