find-skills

Pass

Audited by Gen Agent Trust Hub on Mar 31, 2026

Risk Level: SAFECOMMAND_EXECUTIONREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses the npx skills command to perform searches and manage the lifecycle of other agent skills.- [REMOTE_CODE_EXECUTION]: Instructions are provided for the agent to install third-party code packages using npx skills add. While this involves executing remote code, the skill explicitly mandates a verification step to check installation counts and source reputation.- [EXTERNAL_DOWNLOADS]: The skill facilitates downloading configuration and code from external sources, specifically the skills.sh registry and GitHub repositories.- [PROMPT_INJECTION]: The skill processes untrusted search results from an external registry, which represents a surface for indirect prompt injection.
  • Ingestion points: Output from npx skills find [query] is processed by the agent to make recommendations (SKILL.md).
  • Boundary markers: No explicit delimiters are specified for the search result output.
  • Capability inventory: Ability to execute shell commands (npx skills add) which can install and run external code.
  • Sanitization: Step 4 provides a mandatory manual verification protocol for the agent to validate the trustworthiness of external sources based on install metrics and vendor reputation.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 31, 2026, 04:38 AM