hh-mcp
Pass
Audited by Gen Agent Trust Hub on Sep 11, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill uses the
uvxtool to executehh-mcp-ru, which downloads the package from a standard public repository at runtime. - [INDIRECT_PROMPT_INJECTION]: The skill processes data from the external hh.ru API, which could potentially contain malicious instructions or unexpected content from third-party sources (like candidate resumes or job descriptions).
- Ingestion points: Results from API methods such as
hh_search_methodsandhh_call_method(e.g., job descriptions, candidate messages). - Boundary markers: None explicitly defined in the instructions for separating untrusted API data from prompt instructions.
- Capability inventory: Network access to the HeadHunter API, file system access as needed by the MCP server, and the ability to read and write data (writing requires explicit user confirmation).
- Sanitization: Relies on the underlying LLM's internal safety guardrails when processing external text content.
Audit Metadata