deepgram-python-text-intelligence

Pass

Audited by Gen Agent Trust Hub on May 12, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill includes instructions to download and install additional components using npx skills add deepgram/skills. This is a standard mechanism for extending functionality within the vendor's ecosystem.
  • [PROMPT_INJECTION]: The skill is designed to process untrusted data from external text inputs and hosted URLs via the client.read.v1.text.analyze method. This creates a surface for indirect prompt injection where malicious instructions embedded in the analyzed text could attempt to influence the agent's behavior.
  • Ingestion points: Untrusted data enters the context through the text and url fields in the request parameter of the analyze function in SKILL.md.
  • Boundary markers: The provided examples do not demonstrate the use of delimiters or specific instructions to the model to ignore embedded commands in the source text.
  • Capability inventory: The skill uses the Deepgram SDK to perform network requests for sentiment, summarization, and topic detection.
  • Sanitization: There is no explicit sanitization or filtering of the input text shown in the quick start examples.
Audit Metadata
Risk Level
SAFE
Analyzed
May 12, 2026, 01:46 PM