tavily-search
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill communicates with the vendor's API endpoint at
api.aisa.oneto retrieve search results. This is expected behavior for a search utility and involves communication with the service provider's infrastructure. - [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external websites via search snippets and summaries.
- Ingestion points: External content is retrieved in
scripts/search_client.pyand printed to the standard output. - Boundary markers: The output uses clear visual delimiters (e.g.,
=======) and structured labels to separate results. - Capability inventory: The script is a standalone Python client using only the standard library; it lacks capabilities for file system modification, privilege escalation, or arbitrary command execution.
- Sanitization: The script performs basic content truncation (e.g.,
snippet[:200]) before displaying information to the agent.
Audit Metadata