find-mcp-directories
Pass
Audited by Gen Agent Trust Hub on Jun 17, 2026
Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes shell-based curl commands to interact with its primary API. This functionality is accompanied by detailed instructions on using subshells and local environment files to protect API keys from exposure to the model context.
- [DATA_EXFILTRATION]: The skill initiates network requests to api.servicegraph.co to fetch directory data. This domain is consistent with the service's purpose as defined in the skill metadata.
- [PROMPT_INJECTION]: The skill ingests and processes data from an external API, creating a surface for indirect prompt injection. 1. Ingestion points: JSON responses from api.servicegraph.co processed in SKILL.md. 2. Boundary markers: Not defined for the processed output. 3. Capability inventory: Shell execution capability (via curl). 4. Sanitization: No explicit sanitization or validation of the API-provided data is specified.
Audit Metadata