Lead Research Assistant
Pass
Audited by Gen Agent Trust Hub on Jun 3, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: No security issues detected. The skill provides helpful templates and guidance for sales research without any suspicious behaviors.\n- [NO_CODE]: The skill does not include any scripts or executable logic, eliminating risks associated with code execution or dependency vulnerabilities.\n- [PROMPT_INJECTION]: Evaluated for Indirect Prompt Injection. The skill is designed to process external information about companies and contacts. While this involves processing untrusted data, it is the primary purpose of the skill and is handled via structured templates.\n
- Ingestion points: Data retrieved from external web searches or provided lead lists (referenced in SKILL.md).\n
- Boundary markers: Templates use clear Markdown headers and tables to structure external data.\n
- Capability inventory: Uses office-mcp tools create_docx (file writing) and xlsx_to_json (file reading).\n
- Sanitization: No specific sanitization instructions are included; the skill relies on model-level content safety filters.
Audit Metadata