exa
Pass
Audited by Gen Agent Trust Hub on Mar 2, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill package is entirely composed of markdown documentation, configuration metadata, and reference guides. No executable code such as Python scripts, JavaScript files, or shell scripts is included.
- [NO_CODE]: The package contains no implementation logic; it serves purely as an instructional layer to guide an AI agent in using externally configured MCP tools.
- [SAFE]: All external URLs and service references point to the official Exa.ai documentation and the author's consistent repository on GitHub (jmagar/claude-homelab).
- [SAFE]: Prompt instructions regarding skill invocation are designed for functional reliability and do not attempt to override safety filters, exfiltrate user data, or hide malicious intent.
Audit Metadata