groq-inference

Pass

Audited by Gen Agent Trust Hub on Mar 22, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSDATA_EXFILTRATION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill facilitates the installation and use of official AI SDKs and utility libraries including 'groq', 'anthropic', 'tenacity', and others such as 'deepgram-sdk' and 'elevenlabs'.
  • [DATA_EXFILTRATION]: Legitimate network communication is established with Groq and other reputable AI service providers to fulfill inference requests.
  • [DATA_EXFILTRATION]: The skill provides patterns for reading local audio and image files to transmit them to Groq's API for transcription or analysis, which is consistent with its intended functionality.
  • [COMMAND_EXECUTION]: Documentation mentions a server-side code execution capability within specific Groq models; this is an inherent feature of the remote API and not a local code execution vulnerability within the skill itself.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 22, 2026, 06:50 PM