pr-improver
An agent skill by trailofbits, from trailofbits/skills. Tags: automation, code-quality, debugging, developer-tools, github.
What it does
Runs an autonomous review-and-fix improvement loop over the current branch's changes until a PR review comes back clean, scoped mechanically to the directories the branch touched. Reviews are performed by an installed PR-review skill (default: pr-review-toolkit's review-pr). Use to fix review findings on a branch before opening or updating a pull request ('clean up this branch', 'fix this PR until review passes', 'run review-and-fix on my changes'). NOT for a one-time review — run the PR-review skill directly.
Install
With the skills CLI, which installs into Claude Code, Codex, Cursor and other agents:
npx skills add trailofbits/skills --skill pr-improverOr copy the skill folder into Claude Code's skills directory by hand (~/.claude/skills for every project, or .claude/skills inside one):
git clone --depth 1 https://github.com/trailofbits/skills
cp -r skills/plugins/code-improver/skills/pr-improver ~/.claude/skills/pr-improverSafety box score
Grade C, moderate risk: 32 of 100 risk points, rated 2026-10-19. The grade rates what the skill and its scripts can reach on the machine of whoever installs it, across eight categories from shell execution to secrets access. How the score works.
This skill runs an autonomous review/fix loop over the current branch and can repeatedly edit the working tree until a PR review is clean. Its main risks are broad autonomy, shell/workflow execution, and reaching into home dot-directories to locate an installed workflow implementation.
| Category | Level | Why |
|---|---|---|
| Code & shell execution | 2 (moderate) | It explicitly uses `Bash` and `Workflow` to launch code, including `workflows/improve.js` via `scriptPath`, so it can execute scripts and shell commands. |
| Network access | 0 (none) | The skill text does not download from or send data to remote hosts; it only inspects the repo and locates a local workflow. |
| Filesystem reach | 2 (moderate) | It searches outside the repo in home dot-directories with `find ~/.claude ~/.codex . -maxdepth 7 -path '*/code-improver/workflows/improve.js'`, so it touches user files beyond the project. |
| Secrets & credentials | 0 (none) | There is no instruction to read tokens, keys, passwords, `.env` files, or other secret stores. |
| Privilege & persistence | 0 (none) | It does not request sudo, global installs, services, shell startup edits, or other privilege-changing actions. |
| Instruction hijack surface | 0 (none) | The file does not tell the agent to ignore instructions, hide behavior, or follow fetched content; it mostly constrains the loop behavior. |
| Irreversible actions | 3 (high) | It is designed to "loop a PR reviewer and a fixer subagent" until the review is clean, and says the run "needs no babysitting" and can only complete on success or escalation. |
| Transparency | 2 (moderate) | Core behavior depends on an external dynamic workflow, `workflows/improve.js`, which is not included in the reviewed file and may be resolved from installed locations. |
Before installing
- Read the installed `workflows/improve.js` and the reviewer skill it calls; this file only references them.
- Try it on a throwaway branch first and watch the working tree for unexpected edits.
- Confirm you are comfortable with an autonomous loop that can keep changing files until review is clean.
Rated 2026-10-19 by an automated review of 1 file from the skill's repository.
Source
- Repository
- trailofbits/skills (all skills from this repository)
- Path
- plugins/code-improver/skills/pr-improver/SKILL.md
- Branch
- main
- Collection
- code-improver
- Updated
- 2026-09-19
Related skills
- code-improver — Runs an autonomous review-and-fix improvement loop over any code target — a skill, plugin, module, or directory — using a reviewer the user names.
- skill-improver — Runs an autonomous review-and-fix improvement loop over a Claude Code skill until a review comes back clean, with a cross-round findings ledger.
- c-review — Performs comprehensive C/C++ security review for memory corruption, integer overflows, race conditions, and platform-specific vulnerabilities.
- dwarf-expert — Analyzes DWARF debug information in compiled binaries.
- github-triage — Triages a repository's open GitHub issues and pull requests via the gh CLI.
- modern-cpp — Guides C++ code toward modern idioms (C++20/23/26). Use when writing new C++ code, modernizing legacy patterns, or working on security-critical C++.