skills/sarvamai/skills/sarvam-mcp/Gen Agent Trust Hub

sarvam-mcp

Pass

Audited by Gen Agent Trust Hub on Jul 29, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill references the installation of the sarvam-mcp package through standard Python tool runners like uvx and pip.
  • [COMMAND_EXECUTION]: Provides configuration templates for setting up the MCP server across multiple IDEs and clients, involving standard shell execution of the vendor-supplied tool.
  • [CREDENTIALS_UNSAFE]: Outlines legitimate authentication methods using the SARVAM_API_KEY environment variable or a local ~/.sarvam/credentials file. No hardcoded credentials or unsafe secret handling practices were found.
  • [DATA_EXFILTRATION]: Performs network operations solely to the official Sarvam API endpoint (api.sarvam.ai) as required for its speech and translation capabilities.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 29, 2026, 05:17 AM
Security Audit — agent-trust-hub — sarvam-mcp