skillgate
Trust infrastructure for AI agent skills: static security scanning, machine-checkable Trust Score, dependency checks, report generation (terminal/JSON/SARIF/JUnit/Markdown/HTML), optional skill benchmarking, and registry watch. Bundles a companion medical-compliance skill library (ISO 13485 QMS, IEC 62304 SDLC, IEC 62366 usability, IEC 60601 electrical safety, IEC 81001 cybersecurity, ISO 10993 biocompatibility, IFU/labeling, PMS, CER, change control, dataset/algorithm checks) — these are documented companion skills, not hidden functionality. Use when the user asks to audit, review, or check an AI agent skill's safety, quality, or trustworthiness.
Security Findings
Credential harvest with network exfiltration
MaliciousAnalyzerCredential/secret over-collection exfiltrated via http.client in skillgate.py
Low-provenance referenced GitHub repo
SuspiciousAnalyzerReferenced repo zhaoxinghua09-cell/skillgate is a 0-star repo used as a download source: https://github.com/zhaoxinghua09-cell/skillgate.git
Versions
1- Scanned Nov 5, 2026, 12:11 PMHigh