Secure Code Review Agent Skills

Easy-to-use, customizable, high-quality secure code review skills for AI agents.

Cost-effective skills that deliver zero-shot, actionable results, with support for Markdown and JSON output formats.

My other skills:

• Threat Modeling Agent Skills
• Vulnerability Management Agent Skills

Table of Contents

• CWE Secure Code Review
• Security Requirements Analysis
• How to Use

CWE Secure Code Review

• Used to identify weaknesses based on the 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-secure-code-review/SKILL.md

Example: cwe-secure-code-review/examples/cwe_secure_code_review_report.md

Security Requirements Analysis

• Used to extract security requirements from a specification document.
• Scoped to the specification.
• Works well with both lower-end and higher-end LLMs.
• Follow up with STRIDE threat modeling to gain further insight into threats.

Skill: security-requirements-analysis/SKILL.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 secure code review and save the output to "cwe_secure_code_review_report.md".

Advanced prompt:

- Perform CWE secure code review and save the output to "cwe_secure_code_review_report.md".
- Convert "cwe_secure_code_review_report.md" to "cwe_secure_code_review_report.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.