tavily-search-free
Pass
Audited by Gen Agent Trust Hub on Jul 9, 2026
Risk Level: SAFENO_CODEPROMPT_INJECTION
Full Analysis
- [NO_CODE]: The skill definition in
SKILL.mdreferences a primary execution script located atscripts/tavily_search.py, but this file is missing from the provided skill package. As a result, the actual implementation logic, including how it handles API responses and shell arguments, cannot be verified. - [PROMPT_INJECTION]: The skill facilitates an indirect prompt injection surface by ingesting untrusted data from the web via the Tavily Search API.
- Ingestion points: Search results (titles, content, and URLs) are fetched from the external web and injected into the agent's context in
SKILL.mdvia the expected output oftavily_search.py. - Boundary markers: The instructions do not define boundary markers (e.g., XML tags or delimiters) or provide system instructions to ignore potential commands embedded within the search results.
- Capability inventory: The skill utilizes subprocess execution of Python scripts and interacts with a remote search API.
- Sanitization: There is no evidence of sanitization or filtering applied to the external content before it is processed by the agent.
Audit Metadata