Lead Research Assistant
Pass
Audited by Gen Agent Trust Hub on Oct 23, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill consists of instructional content and output templates designed to guide the AI in performing research tasks using external web sources.
- [SAFE]: No malicious patterns, such as prompt injection, hardcoded credentials, or unauthorized network calls, were found in the skill metadata or instructions.
- [INDIRECT_PROMPT_INJECTION]: The skill's primary function involves processing untrusted external data (e.g., news and LinkedIn profiles), which is an inherent attack surface for indirect prompt injection. Ingestion points include web search results and profile data mentioned in the research areas. Boundary markers are absent in the templates, relying on Markdown structure. The capability inventory is limited to create_docx and xlsx_to_json, which are low-risk tools. No specific sanitization logic is implemented in the skill instructions.
Audit Metadata