NYC

server-skills

Fail

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: CRITICAL
Full Analysis
  • [SAFE] (SAFE): No malicious behaviors detected across all files. The skill serves as a documentation and auditing resource for developers.
  • [EXTERNAL_DOWNLOADS] (SAFE): No remote scripts, package installations, or external downloads are initiated by the skill.
  • [DATA_EXFILTRATION] (SAFE): No access to sensitive files (e.g., SSH keys, AWS credentials) or network exfiltration patterns were found.
  • [PROMPT_INJECTION] (SAFE): No jailbreak attempts, role-play injections, or instructions to ignore system safety protocols are present.
  • [INDIRECT_PROMPT_INJECTION] (LOW): The skill defines patterns for analyzing local source code using Grep. While it ingests file content (ingestion point), it includes specific defensive patterns for path traversal (Checklist 5 in fastapi.md) and relies on standard developer tools (Bash/Grep) for static analysis.
Recommendations
  • Contains 1 malicious URL(s) - DO NOT USE
Audit Metadata
Risk Level
CRITICAL
Analyzed
Feb 17, 2026, 06:04 PM