find-skills
Pass
Audited by Gen Agent Trust Hub on Jun 16, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides instructions for executing the
npx skillsCLI tool to perform various tasks including searching for skills, checking for updates, and managing installations.\n- [EXTERNAL_DOWNLOADS]: The skill facilitates the download and installation of modular code packages from external sources like GitHub and the skills.sh registry.\n- [EXTERNAL_DOWNLOADS]: Mentions fetching and installing guidelines and tools from well-known and reputable organizations such as Vercel Labs and Anthropic.\n- [PROMPT_INJECTION]: The skill searches an external registry (skills.sh) for third-party skills, creating an attack surface where malicious skill metadata could potentially influence agent behavior (Indirect Prompt Injection).\n - Ingestion points: Search results from the
npx skills findcommand enter the agent's context.\n - Boundary markers: None provided in the search output processing instructions.\n
- Capability inventory: Subprocess execution capability through the
npx skills addcommand.\n - Sanitization: Instructions rely on manual verification steps rather than automated sanitization of search results.
Audit Metadata