find-skills
Pass
Audited by Gen Agent Trust Hub on Oct 19, 2026
Risk Level: SAFECOMMAND_EXECUTIONREMOTE_CODE_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses the shell to run
npx skillscommands for searching, adding, and updating skills. - [REMOTE_CODE_EXECUTION]: The skill's primary function is to install and execute code from remote GitHub repositories or other sources via the
npx skills addcommand. While the instructions recommend using trusted sources, the command allows installation of any package. - [INDIRECT_PROMPT_INJECTION]: The skill processes external data from CLI search results and the
skills.shwebsite. This data (such as skill descriptions or names) could contain malicious instructions designed to manipulate the agent's behavior. - Ingestion points: Search results from
npx skills find [query]and data from theskills.shleaderboard. - Boundary markers: None identified. The instructions do not specify the use of delimiters when processing external search results.
- Capability inventory: The agent can execute code through package installation (
npx skills add) and initialize new projects (npx skills init). - Sanitization: The skill includes a manual verification step ('Step 4: Verify Quality Before Recommending') instructing the agent to check install counts and source reputation, which serves as a heuristic mitigation.
Audit Metadata