find-skills
Warn
Audited by Gen Agent Trust Hub on Apr 13, 2026
Risk Level: MEDIUMEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill is designed to fetch and install external packages from GitHub using commands like 'skilllite add owner/repo@skill'.\n- [COMMAND_EXECUTION]: Employs the 'run_command' capability to execute shell-based search and installation routines ('npx skills find', 'skilllite add').\n- [REMOTE_CODE_EXECUTION]: By installing new skills, the agent effectively integrates and executes remote code or instructions from third-party repositories into its operational environment.\n- [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection via untrusted search results.\n
- Ingestion points: Results from the 'npx skills find' command (SKILL.md).\n
- Boundary markers: No delimiters or warnings are used to isolate search result content.\n
- Capability inventory: The 'run_command' tool is available to execute the 'skilllite add' command based on findings.\n
- Sanitization: No evidence of validation or filtering for external registry metadata.\n- [EXTERNAL_DOWNLOADS]: Fetches guidelines and skills from Vercel Labs' official GitHub repository.
Audit Metadata