skills/jsoyer/dotfiles/research/Gen Agent Trust Hub

research

Pass

Audited by Gen Agent Trust Hub on Jun 19, 2026

Risk Level: SAFECREDENTIALS_UNSAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [CREDENTIALS_UNSAFE]: The script reads authentication tokens from the local ~/.mcp-auth/ directory. It includes a validation function that parses the JWT (JSON Web Token) to ensure it was issued by the official Tavily endpoint and has not expired before attempting to use it.
  • [EXTERNAL_DOWNLOADS]: When no local credentials are found, the skill uses npx to download and run the mcp-remote utility. This tool is a standard part of the Model Context Protocol ecosystem used to facilitate secure remote authentication.
  • [COMMAND_EXECUTION]: The script executes several system commands including find to locate configuration files, curl to communicate with the Tavily research API, and jq to process and validate JSON data structures.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 19, 2026, 07:40 PM
Security Audit — agent-trust-hub — research