muapi-director

Pass

Audited by Gen Agent Trust Hub on Sep 3, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external inputs such as user-provided topics and audio transcripts (generated via Whisper ASR) and interpolates them into prompt templates for image and video generation. There is a lack of boundary markers or sanitization in the prompt construction, which allows external content to potentially influence the behavior of the downstream generation models. For example, the narration and topic fields from beats.json are used directly in prompts within aroll_clips.py and clips.py.
  • [COMMAND_EXECUTION]: The orchestration scripts rely heavily on subprocess.run to execute system binaries, specifically ffmpeg, ffprobe, and curl. These are used for video assembly, media metadata probing, and file transfers between the local system and remote services. Evidence includes numerous subprocess.run calls found in assemble.py, aroll_assemble.py, asr_beats.py, and muapi_client.py.
  • [EXTERNAL_DOWNLOADS]: The skill communicates with well-known service endpoints including api.muapi.ai and api.openai.com to perform AI generation and transcription tasks. It uses curl to upload local media assets to these services and download the resulting media files.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 3, 2026, 08:51 AM