text-to-speech

Pass

Audited by Gen Agent Trust Hub on Mar 10, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill requires standard packages including the 'edge-tts' Python package and Node.js packages '@modelcontextprotocol/sdk', 'ws', and 'fs-extra'. These are reputable libraries for the described speech synthesis and MCP server tasks.
  • [DATA_EXFILTRATION]: Text content is sent to the Microsoft Edge TTS endpoint at 'speech.platform.bing.com' for audio generation. This is a well-known service from a trusted provider and the data transmission is necessary for the skill's primary function.
  • [PROMPT_INJECTION]: The skill processes untrusted documents and includes features to summarize them via an LLM, creating a vulnerability surface for indirect prompt injection.
  • Ingestion points: Active VS Code editor text and markdown documents provided by the user.
  • Boundary markers: No specific boundary markers or 'ignore embedded instructions' warnings are documented for the summarization or synthesis prompts.
  • Capability inventory: The skill can read local files, establish network WebSocket connections, write audio files to the disk, and interact with the VS Code Language Model API.
  • Sanitization: While the skill strips markdown formatting for auditory clarity, it does not mention content validation or sanitization to prevent the execution of instructions embedded in the documents.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 10, 2026, 03:54 AM