speak-hello-world

Pass

Audited by Gen Agent Trust Hub on Sep 9, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill follows security best practices by managing API credentials through environment variables (SPEAK_API_KEY, SPEAK_APP_ID) instead of hardcoding sensitive data.
  • [SAFE]: The libraries utilized (@speak/language-sdk and speak_sdk) are appropriate for the service described and appear to be legitimate vendor resources.
  • [SAFE]: External references are limited to reputable documentation and service sites (speak.com and openai.com).
  • [INDIRECT_PROMPT_INJECTION]: The skill processes user-submitted audio and text for tutoring purposes, which is a potential surface for indirect injection.
  • Ingestion points: External data enters via client.sendTurn, client.assessPronunciation, and session.submitResponse within the TypeScript and Python examples.
  • Boundary markers: None are explicitly used in the provided sample code to delimit user input from instructions.
  • Capability inventory: The skill performs network requests to the Speak API and accesses local files for audio recording processing.
  • Sanitization: The skill relies on the underlying SDKs and backend service for data validation and sanitization.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 9, 2026, 03:48 AM
Security Audit — agent-trust-hub — speak-hello-world