find-skills
Pass
Audited by Gen Agent Trust Hub on Oct 26, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill relies on the
npx skillsCLI to search for, install, and update agent skills. This involves executing shell commands to interact with the package manager. - [INDIRECT_PROMPT_INJECTION]: The skill retrieves and displays metadata (names, descriptions) from an external registry (
skills.sh) based on user search queries. - Ingestion points: Search results from
npx skills findinSKILL.md. - Boundary markers: None present; external content is presented directly to the user/agent.
- Capability inventory: Capability to install external packages globally with skipped confirmations (
npx skills add -g -y). - Sanitization: No explicit sanitization or validation of the fetched skill metadata is described in the prompt instructions.
Audit Metadata