tavily-literature-review
Pass
Audited by Gen Agent Trust Hub on Nov 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill fetches and summarizes data from multiple external academic and publisher websites, which presents an attack surface where malicious content on those sites could influence the agent's behavior. \n
- Ingestion points: External content is fetched from domains like arxiv.org, pubmed.ncbi.nlm.nih.gov, and nature.com in
SKILL.mdvia thetvlytool.\n - Boundary markers: No explicit delimiters or instructions to ignore embedded commands are present in
SKILL.mdfor the fetched content.\n - Capability inventory: The agent can search, extract data, and write files to the local
reviews/directory as specified inSKILL.md.\n - Sanitization: No evidence of sanitization or validation of the external content is provided before the agent processes the text.\n- [EXTERNAL_DOWNLOADS]: The skill includes instructions to download a necessary dependency from the vendor's repository.\n
- Evidence:
SKILL.mdcontains an instruction to runnpx skills add tavily-ai/skills --skill tavily-searchif the dependency is missing.
Audit Metadata