check-malicious-code

To audit installed skills for malicious code, you run a full security scan that detects hidden instructions and vulnerabilities across all official, example, and user directories. It produces a structured report prioritizing the identified risks.

Pattern-based detection uses regex to scan skill files like SKILL.md, scripts, and documentation for malicious code. It applies context-aware analysis with line-level context and a predefined whitelist to accurately identify vulnerabilities.

Verify newly installed skills for safety by executing a security audit before enabling them. This scans for malicious code and hidden instructions across your skill files, generating Markdown and JSON reports to review before integration.

Yes, the malicious code audit supports a whitelist for safe patterns. It uses predefined pattern sets alongside contextual analysis to filter out safe code, ensuring the structured report only highlights genuine vulnerabilities.

Yes, you can get a JSON report of security vulnerabilities found in installed skills. The audit generates comprehensive reports in both Markdown and JSON formats, prioritizing the identified vulnerabilities for structured review.