assemblyai

Pass

Audited by Gen Agent Trust Hub on Jun 21, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill installs the official assemblyai SDK and well-known helper libraries (python-dotenv, pyaudio) via pip to facilitate audio processing and secret management.
  • [COMMAND_EXECUTION]: The setup instructions involve standard package installation commands and environment variable configuration in the shell.
  • [PROMPT_INJECTION]: The skill presents an indirect prompt injection surface as it ingests transcripts from external audio sources and processes them via the LeMUR LLM layer (e.g., in lemur_qa and lemur_summarize).
  • Ingestion points: Transcription of external audio files/URLs via the aai.Transcriber client.
  • Boundary markers: No specific delimiters or safety instructions are used to separate transcript content from LeMUR tasks in the provided examples.
  • Capability inventory: The skill performs network requests to the AssemblyAI API and utilizes LLM reasoning over transcript data.
  • Sanitization: No explicit sanitization or filtering of transcript text is performed before it is passed to the LLM layer.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 21, 2026, 07:31 PM
Security Audit — agent-trust-hub — assemblyai