hiring-signal-outreach
Pass
Audited by Gen Agent Trust Hub on Oct 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill retrieves and processes external job descriptions and candidate details to create personalized outreach. This creates a risk where malicious instructions embedded in a public job posting could influence the agent's behavior during the drafting phase.\n
- Ingestion points: Job descriptions, candidate profiles, and role summaries fetched from external sites like LinkedIn, Indeed, and Apollo (Step 1 and Step 3).\n
- Boundary markers: The skill does not define clear boundaries or instructions to treat retrieved external content as untrusted data.\n
- Capability inventory: The skill has access to
web-search,job-search, andemail-draftingcapabilities, and it stores configuration data in the local file system (clients/directory).\n - Sanitization: There is no evidence of sanitization or filtering logic applied to the external text before it is used to guide the LLM's email generation process.\n- [EXTERNAL_DOWNLOADS]: The skill metadata includes an installation instruction using
npx, which is the standard package execution method for this platform and is used here for installation purposes.
Audit Metadata