transcribe

Pass

Audited by Gen Agent Trust Hub on Oct 23, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes transcription data derived from untrusted external audio and video files. The agent is subsequently instructed to perform semantic refinement on this text, which creates an attack surface where malicious instructions embedded in the audio could influence the agent's behavior.
  • Ingestion points: The agent reads transcripts generated from user-provided media files via the scripts/transcribe.ts utility.
  • Boundary markers: The instructions lack explicit delimiters or warnings to ignore embedded commands within the transcription results during the refinement phase.
  • Capability inventory: The skill has the ability to read and write files and perform network operations.
  • Sanitization: No sanitization or filtering is applied to the transcribed text before it is presented to the agent for processing.
  • [EXTERNAL_DOWNLOADS]: The scripts/transcribe.ts script performs network operations to api.elevenlabs.io to transmit audio data and receive transcripts. This targets a well-known voice AI service and is a core part of the skill's intended functionality.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 23, 2026, 02:38 PM