web-research
Pass
Audited by Gen Agent Trust Hub on Mar 10, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection.\n
- Ingestion points: Untrusted data enters the agent context through the
WebSearchandWebFetchtools as described in Step 2 and Step 3 of SKILL.md.\n - Boundary markers: The skill instructions do not specify the use of delimiters or 'ignore embedded instructions' warnings when processing content fetched from the web.\n
- Capability inventory: The agent possesses
Write,TaskCreate, andWebSearchcapabilities, which could allow it to perform unintended actions or modify local files based on malicious content found in search results.\n - Sanitization: There is no evidence of sanitization, escaping, or validation of the external content before it is triaged or used to build the source map.
Audit Metadata