perplexity

Pass

Audited by Gen Agent Trust Hub on May 19, 2026

Risk Level: SAFE
Full Analysis
  • [CREDENTIALS_UNSAFE]: The skill follows security best practices by requiring the PERPLEXITY_API_KEY to be provided as an environment variable instead of hardcoding any secrets in the source code.
  • [DATA_EXFILTRATION]: All network requests in scripts/search.mjs are directed to api.perplexity.ai. This is the official, well-known domain for the Perplexity service and no sensitive local data is sent to the endpoint.
  • [EXTERNAL_DOWNLOADS]: The skill does not perform any external package installations (npm/pip) or remote script executions. It uses standard Node.js functionality.
  • [COMMAND_EXECUTION]: The script scripts/search.mjs processes user-supplied query strings into an API request but does not pass this input into shell execution or other dangerous functions.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external search results. While this is a surface for indirect injection, the script only performs text formatting and lacks the capabilities (like file writing or code execution) required for such an attack to be successful.
Audit Metadata
Risk Level
SAFE
Analyzed
May 19, 2026, 10:47 AM
Security Audit — agent-trust-hub — perplexity