deepgram-java-text-intelligence

Pass

Audited by Gen Agent Trust Hub on Jul 3, 2026

Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [PROMPT_INJECTION]: The skill processes untrusted text input via the ReadV1RequestText builder. This creates a surface for indirect prompt injection if the text contains instructions meant to override agent behavior. The current documentation lacks boundary markers (like specific delimiters) or instructions to treat the analyzed text as passive data.
  • [EXTERNAL_DOWNLOADS]: Instructions are provided to install further capabilities using npx skills add deepgram/skills. This involves fetching and executing code from an external vendor-maintained repository.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 3, 2026, 09:56 AM
Security Audit — agent-trust-hub — deepgram-java-text-intelligence