find-skills
Pass
Audited by Gen Agent Trust Hub on Apr 6, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill facilitates the discovery and download of external agent packages from the skills.sh registry and GitHub repositories. It encourages the use of established sources such as Vercel Labs' official repositories.
- [REMOTE_CODE_EXECUTION]: Instructions provide for the automated installation of remote code via the 'npx skills add' command. The documentation suggests using the '-y' flag to bypass confirmation prompts during the installation process.
- [COMMAND_EXECUTION]: The agent is instructed to execute shell commands using the 'npx skills' CLI tool for searching, checking, and updating capabilities.
- [PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection and command injection. User-provided queries are interpolated directly into shell commands ('npx skills find [query]') without explicit sanitization instructions or boundary markers. Ingestion point: SKILL.md; Boundary markers: Absent; Capability inventory: Shell execution via npx (SKILL.md); Sanitization: Absent.
Audit Metadata