talking-head-recut

Pass

Audited by Gen Agent Trust Hub on Sep 13, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructs the agent to download and run the hyperframes package from the NPM registry using npx, as well as to update the skill via npx hyperframes skills update.
  • [REMOTE_CODE_EXECUTION]: The use of npx to fetch and execute CLI tools at runtime constitutes remote code execution from an external registry.
  • [DYNAMIC_EXECUTION]: The skill involves the dynamic generation of HTML fragments and JavaScript code (GSAP animations). This code is later rendered into video frames using a headless browser engine, creating a runtime execution surface for generated content.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests transcript.json data, which is generated from external audio/video files. This content is used to drive the text and timing of the generated graphic cards, presenting a surface for indirect prompt injection if the source media contains malicious instructions designed to influence the agent's HTML/JS generation.
  • [COMMAND_EXECUTION]: The workflow relies on multiple shell commands for media processing, including ffmpeg for audio extraction and re-encoding, ffprobe for metadata retrieval, and npx for transcription and rendering tasks.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 13, 2026, 03:57 PM
Security Audit — agent-trust-hub — talking-head-recut