plugin-security-checker

Security scanning for Claude Code plugins identifies vulnerabilities, code obfuscation, and insecure patterns across manifests, Python/JavaScript code, and agents. Run the plugin-security-checker on your plugin directory to generate JSON, HTML, and Markdown reports with detailed findings and risk scores.

Plugin scanners detect 91 specialized vulnerability patterns including credential leaks, obfuscation techniques, anti-patterns in Python and JavaScript, unsafe API usage, and configuration weaknesses. Results map findings to MITRE ATT&CK and ATLAS frameworks for threat context.

Yes. Plugin-security-checker enriches findings with MITRE ATT&CK and ATLAS mappings, STIX bundles, and cross-framework risk scoring to contextualize vulnerabilities and support compliance audits and pre-install reviews.

Export plugin security analysis results as JSON, Markdown, or HTML reports with MITRE ATT&CK and ATLAS threat intelligence exports. Auditable outputs support pre-install reviews, security audits, and risk documentation.

Plugin security scanning enforces a 500MB memory limit with Bloom filter and Trie validation for near-zero false positives. Adaptive routing by file type optimizes performance across manifests, scripts, and agent configurations.

Consensus voting across 91 specialized pattern agents reduces false positives and ensures high-confidence vulnerability detection. Combined with Bloom filter and Trie validation, this approach delivers auditable security findings with minimal noise.