drug-discovery-search

Pass

Audited by Gen Agent Trust Hub on Apr 16, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes shell commands to interact with its internal logic and resolve file paths.
  • It instructs the agent to use find to locate the search script within the environment's plugin cache (~/.claude/plugins/cache).
  • It uses a bash wrapper (scripts/search) to execute search queries and handle API key configuration.
  • [EXTERNAL_DOWNLOADS]: The skill interacts with the Valyu platform for its primary functionality.
  • It makes network requests to https://api.valyu.ai/v1/search using the Node.js built-in fetch API.
  • Documentation references external dependencies valyu (PyPI) and valyu-js (NPM) for developers looking to integrate the service into their own projects.
  • [CREDENTIALS_UNSAFE]: The skill handles a Valyu API key to authenticate requests.
  • It provides a setup command scripts/search setup <api-key> to store the user-provided key. This is a standard practice for local tool configuration and does not involve hardcoded secrets or exfiltration to unauthorized third parties.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 16, 2026, 08:25 AM