aisa-tavily
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill possesses an ingestion surface for untrusted external data retrieved from the web during search and extraction tasks. This is a standard characteristic of research-oriented skills.\n
- Ingestion points:
scripts/extract.mjsandscripts/search.mjsfetch and output content from external URLs to the agent context.\n - Boundary markers: The output uses Markdown headings and horizontal rules (---) to separate external content from system instructions.\n
- Capability inventory: The agent reads the extracted text to fulfill user research requests.\n
- Sanitization: The skill prints the raw content retrieved from the API relay without additional filtering or sanitization.
Audit Metadata