code-review-skills

Name: code-review-skills
Author: akkie76

by akkie76

Verified

Evidence-based code review skills for Codex and Claude Code

85stars

0forks

Python

Added 9/28/2026

Installation

# Add to your Claude Code skills
git clone https://github.com/akkie76/code-review-skills

Getting Started

Guides for using ai agents skills like code-review-skills.

Caveman: Cut Claude Token Use by 65%
How agent-side prompt compression works, when to use it, and when not to.
How to Test an AI Skill: A Practical Guide to Evals
Evaluate skills with realistic fixtures, a baseline, and repeatable checks.
Getting Started with AI Skills

Security ReportVerified

Last scanned: 9/29/2026

{
  "issues": [],
  "status": "PASSED",
  "scannedAt": "2026-09-29T10:25:55.958Z",
  "npmAuditRan": true,
  "pipAuditRan": true,
  "promptInjectionRan": true
}

Compare with similar skills

See how code-review-skills compares with popular alternatives.

code-review-skills vs hermes agent

See comparison

code-review-skills vs ecc

See comparison

README.md

Frequently Asked Questions

What is code-review-skills?

code-review-skills is an open-source ai agents skill for AI coding assistants such as Claude Code, Codex CLI, and ChatGPT, built by akkie76. Evidence-based code review skills for Codex and Claude Code. It has 85 GitHub stars.

Is code-review-skills safe to use?

Yes. code-review-skills passed SkillsLLM's automated security scan — a dependency vulnerability audit plus prompt-injection heuristics — with no high-severity issues. You can read the full report in the Security Report section on this page.

How do I install code-review-skills?

Clone the repository with "git clone https://github.com/akkie76/code-review-skills" and add it to your Claude Code skills directory (see the Installation section above).

What programming language is code-review-skills written in?

code-review-skills is primarily written in Python. It is open-source under akkie76 on GitHub, so you can review or fork the full source.

Are there alternatives to code-review-skills?

Yes. SkillsLLM lists many other AI Agents skills you can browse and compare side by side. Open the AI Agents category from the badge at the top of this page, or use the Related Skills and comparison links further down to weigh code-review-skills against similar tools.

AI Courses · Agentic AI, LLM Engineering & MCP

Beginner to professional · First lesson free · Lifetime access

Comments (0)

to leave a comment.

No comments yet. Be the first to share your thoughts!

Related Skills

rea

by morluto

Reverse engineer anything with agents, from app behavior down to native binaries.

456,329

ViralForge glyph-kit

View on GitHub Compare Download ZIP Scan for vulnerabilities

From $49 · 30 days · terms & refunds

⚠️ Third-Party Software Notice

This skill is third-party open-source software developed and hosted independently on GitHub. SkillsLLM is an informational directory and does not control or maintain the underlying repository.

Any security checks, ratings, or warnings displayed by SkillsLLM are automated and limited in scope. They do not constitute a security certification or guarantee that the software is safe, error-free, or free from malicious code, vulnerabilities, compromised dependencies, or prompt-injection risks.

Review the source code, permissions, dependencies, and configuration before installing or running any third-party skill. Use is at your own risk. To the maximum extent permitted by applicable law, SkillsLLM is not liable for losses arising from third-party software.

Read the Terms of Service

Code Review Skills

日本語

An evidence-driven code-review workflow for Codex and Claude Code.

This Skill is based on ideas from コードレビューの教科書 and independently adapts them for AI coding agents. It does not reproduce or replace the book.

Version: v1.0.0 · License: MIT

The documented review-output format is stable from 1.0.0. Review decisions are still model- and context-dependent: this Skill does not guarantee that every defect will be found or that repeated reviews will produce identical findings.

Get started

Clone the release tag so the installed package has a known version:

git clone --depth 1 --branch v1.0.0 https://github.com/akkie76/code-review-skills.git
cd code-review-skills

Codex

mkdir -p ~/.codex/skills
cp -R dist/codex/evidence-code-review ~/.codex/skills/

Then start a new Codex task and ask, for example:

Use the evidence-code-review skill to review the changes on this branch against main. Focus on correctness, regressions, and tests. Do not modify files.

Claude Code

mkdir -p ~/.claude/skills
cp -R dist/claude-code/evidence-code-review ~/.claude/skills/

Start a new Claude Code session and invoke the Skill, for example:

/evidence-code-review Review the changes on this branch against main. Focus on correctness, regressions, and tests. Do not modify files.

For project-only installation, Windows instructions, verification, updates, removal, and troubleshooting, see the installation guide.

What it reviews

Establishes the requested scope and repository-specific rules.
Traces changed behavior through affected code paths, not only changed lines.
Checks correctness, interfaces, design, security, reliability, tests, and documentation according to risk.
Reviews code without modifying it unless you separately ask for changes.

What a review looks like

Actionable comments start with an action level and viewpoint:

MUST: a demonstrated issue that must be resolved before merge.
SHOULD: a concrete risk that normally should be addressed, but may be deferred by an explicit decision.
BETTER: an optional alternative with a specific benefit.
NITS: a minor, non-blocking correction.

The viewpoint explains the concern, such as Functionality, Test, Simplicity, or Document. A finding should give its location and concrete evidence, explain when it occurs and its impact, and suggest a proportionate direction.

MUST(Functionality): Advance the page before requesting the next result set

Location: `src/client.ts:42`
When the API returns a full page, this loop requests the same page again because
`page` never changes. That duplicates results and increases request volume;
advance the page before the next request.

This is an illustrative example. See the review comment convention for the full action-level, viewpoint, evidence, and communication rules.

Project-specific context

Languages, frameworks, architecture, and test commands differ by repository. Put the information needed for an accurate review in that repository's AGENTS.md, CLAUDE.md, or a project document referenced by those files. Useful context includes version constraints, lifecycle or concurrency rules, architecture conventions, and verification commands. Do not store project information in the installed package; updates may replace it.

Limits and expectations

Results depend on the model, agent version, available tools, review scope, and project information. A stronger model may help, but does not guarantee a correct or complete review.
Use this Skill alongside tests and human review; it is not a substitute for either and cannot guarantee zero missed defects or false positives.
The Skill does not grant additional permissions. File access and approval behavior are controlled by the host agent and its settings.
Reviewing large changes may consume substantial model tokens.
The latest evaluation report describes a small synthetic fixture campaign, not production review accuracy; see the evaluation guide for context and other reports.

Updates and support

The versioning policy describes the stable output contract and how changes map to versions.
See the installation guide for updating or removing an installation.
Evidence-backed bug and accuracy reports are welcome under the support policy. Pull requests require prior agreement.
Report suspected vulnerabilities privately as described in SECURITY.md.

License

The original material in this repository is licensed under the MIT License.