exa-performance-tuning
Pass
Audited by Gen Agent Trust Hub on Sep 12, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill serves its stated purpose of providing performance optimization patterns for the Exa API without any hidden or malicious functionality.
- [DATA_EXPOSURE_AND_EXFILTRATION]: The provided code snippets follow security best practices by referencing sensitive credentials through environment variables (
process.env.EXA_API_KEY) rather than hardcoding them. - [EXTERNAL_DOWNLOADS]: The skill references standard, well-known Node.js packages (
exa-jsandlru-cache) which are standard dependencies for the described functionality. - [INDIRECT_PROMPT_INJECTION]: While the code handles user-provided queries, it does not involve any pattern where untrusted data is processed by the agent in a way that could lead to instruction override or unauthorized actions.
Audit Metadata