clinical-trials-search
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests and processes data from an external search API (Valyu), creating a potential surface for indirect prompt injection if the source content contains adversarial instructions.
- Ingestion points: Data enters the agent context through the
scripts/search.mjsscript, which fetches trial details, phases, and outcomes fromapi.valyu.ai. - Boundary markers: The skill does not define specific delimiters or instructions for the agent to ignore embedded commands within the search results.
- Capability inventory: The skill's scripts have the ability to read and write to the user's home directory (specifically the
~/.valyuconfiguration directory) and perform network operations. It does not appear to have arbitrary command execution or broad filesystem access capabilities. - Sanitization: The script parses the external JSON response but does not perform sanitization or filtering of the text content before returning it to the agent.
Audit Metadata