Secure Code Review Agent Skills

Easy-to-use, easy-to-customize, and high-quality secure code review skills for AI agents.

Supporting Markdown and JSON output formats.

My other skills:

• Threat Modeling Agent Skills

Table of Contents

• Secure Code Review
  • CWE
• How to Use

Secure Code Review

CWE

• Used to identify weaknesses based on Common Weakness Enumeration (CWE).
• Applies structured, single-step kill chain reasoning without considering business context, objectives, or impact.
• Scoped to the application.
• Works well with both lower-end and higher-end LLMs.

Skill: cwe-code-review/SKILL.md

Example: cwe-code-review/examples/cwe_code_review.md

How to Use

• Copy the contents of the markdown directory into your project's .claude/skills/ directory.
• Alternatively, manually upload each SKILL.md file to your Claude app under Customize -> Skills.

Basic prompt:

Perform CWE code review and save the output to "cwe_code_review.md".

Advanced prompt:

- Perform CWE code review and save the output to "cwe_code_review.md".
- Convert "cwe_code_review.md" to "cwe_code_review.html".
- Ensure the `body` CSS rule includes `width: 100%; max-width: 100%;`.
- Ensure the `td` CSS rule includes `word-break: keep-all;`.
- Add a table of contents.
- Make the non-key-value tables sortable.