skills/airsuphq/machine-skill/machine/Gen Agent Trust Hub

machine

Pass

Audited by Gen Agent Trust Hub on Oct 14, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes local Node.js scripts (record.mjs, shot.mjs, inline.mjs) to automate the production of 3D simulations and media artifacts. These scripts leverage system utilities like FFmpeg for video processing and Playwright for browser-based rendering.
  • [EXTERNAL_DOWNLOADS]: Retrieves the Three.js library and associated modules from the jsDelivr CDN, environmental textures from Poly Haven, and audio tracks from Archive.org. These sources are established and well-known service providers in the developer community.
  • [DYNAMIC_EXECUTION]: Utilizes eval() within the browser context during the review phase (shot.mjs) to programmatically manipulate the simulation state, such as toggling cutaways or camera positions, based on agent-generated logic.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests user descriptions via the /machine command. The generation process is guided by a rigorous engineering workflow involving research and physics modeling, which inherently scopes the agent's behavior and reduces the impact of adversarial instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 14, 2026, 02:20 AM