What does Local Codex Review do?

It runs the same Codex PR review process locally, allowing you to catch issues before pushing changes to the main branch.

How do I install and use this skill?

Ensure you have the Codex CLI installed and authenticated. Then, run the provided bash script to perform a local review.

What are the prerequisites for using this skill?

You need to have Codex CLI version 0.144.1 or higher installed and authenticated, and your base branch should be up to date.

Can I use this skill on any branch?

Yes, you can run the review against any branch by specifying a different base reference when invoking the script.

What output can I expect from the Local Codex Review?

The output is presented in markdown format, detailing findings with severity tags and the corresponding file and line numbers.

Local Codex Review

Name: Local Codex Review
Availability: InStock
Author: windmill-labs

Free

Run CI Codex reviews locally before pushing changes.

by windmill-labs17.5k stars on windmill-labs/windmill

2 views

Updated Aug 10, 2026

Get this skill

Free · Opens the source repo

What Local Codex Review does

Local Codex Review allows developers to run the same Codex PR review process locally that would occur in Continuous Integration (CI) environments. This skill is particularly useful for catching issues before they reach the CI stage, enabling developers to address potential problems in their code before pushing changes. The review process is scoped to include both committed and uncommitted changes, ensuring that all modifications are evaluated against the project's coding standards and guidelines.

The skill operates by executing a script that computes the appropriate base reference for comparison and feeds the necessary context to Codex. It uses the same review policies as the CI process, ensuring consistency in findings. The output is presented in markdown format, clearly indicating findings with severity tags (P0, P1, P2) along with the relevant file and line numbers. This allows developers to quickly identify and address critical issues before they become part of the main codebase.

Local Codex Review is designed for developers who are looking to enhance their code quality and maintain compliance with project standards. It is especially beneficial for teams that rely on Codex for code reviews and want to streamline their workflow by catching issues early in the development process. By integrating this skill into their pre-push routine, developers can save time and reduce the likelihood of introducing errors into the main branch.

To use this skill, developers need to have the Codex CLI installed and authenticated, ensuring they are using a compatible version. The skill requires minimal setup and can be easily invoked through a simple bash command, making it accessible for developers at various skill levels.

When to use it

Use Local Codex Review when working on non-trivial changes that you want to validate against project standards before pushing.

When not to use it

This skill is not suitable for reviewing changes that have already been pushed or for scenarios where CI checks are sufficient without local validation.

What you can build with it

Pre-push Validation

Run Local Codex Review to validate your changes against project standards before pushing to the main branch.

Catch Issues Early

Use this skill to identify potential issues in your code before they reach the CI stage, saving time and effort.

Improve Code Quality

Incorporate Local Codex Review into your workflow to enhance code quality and maintain compliance with coding guidelines.

How to install Local Codex Review

View source

1. Install with the skills CLI

npx skills add windmill-labs/windmill/local-review-codex --agent claude-code

2. Or install it manually

Download the skill folder and drop it into ~/.claude/skills/ for all projects, or .claude/skills/ to scope it to one repo. Restart Claude Code so it picks up the new skill.

Anthropic's agentic coding CLI, and the reference implementation of Agent Skills. Drop a skill folder into ~/.claude/skills and Claude Code loads it automatically whenever a task matches the skill's description. Claude Code docs

Inside SKILL.md

Written by windmill-labs

Local Codex Review (pre-push)

Runs the exact same review Codex performs in CI (.github/workflows/codex-pr-review.yml), but locally and scoped to work you have not pushed yet — so you catch what CI would flag before the PR exists. Use this before git push on a non-trivial change.

Correspondence with CI — identical:

Policy: REVIEW.md (severity triage, public-surface checklist, AGENTS.md compliance, test coverage).
Model: gpt-5.6-sol, model_reasoning_effort="xhigh".
Output: markdown starting with ## Codex Review, findings tagged P0 / P1 / P2 with file:line.

Differences from CI — local-only:

Scope is the current branch vs main at the merge-base, including uncommitted changes (CI reviews a pushed PR diff).
Sandbox is read-only (CI uses danger-full-access on an ephemeral runner). Codex reads the diff and files but cannot modify your working tree.
Fresh context is inherent: codex exec is a separate cold process, so it does not anchor on the current chat session — the same reason local-review insists on a subagent.

Prerequisites

codex CLI >= 0.144.1 installed and authed (codex login or OPENAI_API_KEY). Older CLIs reject gpt-5.6-sol with "requires a newer version of Codex". Upgrade with npm install --global @openai/codex@0.144.1 (may need sudo for a global install). Keep this in sync with the pin in .github/workflows/codex-pr-review.yml.
git fetch the base ref if it's stale, so the merge-base is accurate.

Run

bash .agents/skills/local-review-codex/run.sh          # review vs main (default)
bash .agents/skills/local-review-codex/run.sh <base>   # review vs a different base ref

Invoke with bash (or run the executable directly) — the script needs Bash for set -o pipefail; sh is Dash on Debian/Ubuntu and would fail. If main isn't a local branch (e.g. a fresh single-branch checkout), the runner falls back to origin/main automatically.

The script computes BASE_SHA = git merge-base HEAD <base>, feeds Codex REVIEW.md plus a diff context pointing at git diff <BASE_SHA> (which folds in uncommitted edits), and prints the review. It writes only temp files — nothing lands in the working tree.

Relaying the result

Print the Codex output verbatim. Do not re-summarize or filter it — the value of a cold Codex pass is surfacing what the current session would rationalize away. Then decide with the user whether to address findings before pushing.

For a Claude-native review instead, use local-review (branch-diff-reviewer subagent). This skill is the Codex counterpart; run both for independent perspectives.

Frequently asked questions about Local Codex Review

Similar skills

Quality Playbook Generator

Run comprehensive quality audits on any codebase.

PR Draft Summary

Automate PR summary generation for openai-agents-python.

Final Release Review

Streamline your release candidate audits with ease.

Unit Test Vue Pinia

Efficiently write and review unit tests for Vue 3 applications.

Slang Shader Expert

Optimize and integrate Slang shaders with ease.

Telemetry Standards

Ensure consistent event tracking in Supabase Studio.