remix-imessage-ad-from-sample

Pass

Audited by Gen Agent Trust Hub on Jul 3, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill invokes ffmpeg and ffprobe to process video files, extract metadata, and generate thumbnail images.
  • [COMMAND_EXECUTION]: The skill uses npx playwright to record the browser-based chat animations and node to execute local rendering scripts (record-master.js) with a dynamically configured NODE_PATH for dependency resolution.
  • [EXTERNAL_DOWNLOADS]: The skill fetches ad templates, brand assets, and configuration files from the vendor's (Gooseworks-ai) official infrastructure and ElevenLabs proxies for audio generation.
  • [PROMPT_INJECTION]: The skill processes untrusted external data, such as project details, brand research, and source scripts, which represents a potential surface for indirect prompt injection.
  • Ingestion points: Data is ingested through the get_ad_project, get_ad_template, and get_ad_brand tool calls as described in Phase 0 of the workflow.
  • Boundary markers: No explicit delimiters or boundary markers were identified in the instructions for processing the source thread data.
  • Capability inventory: The skill has capabilities for file writing (write_file), local shell command execution (node, ffmpeg, playwright), and network operations for asset uploads.
  • Sanitization: The skill lacks explicit sanitization or validation steps for external text inputs before they are incorporated into the rendering logic or script generation.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 3, 2026, 12:55 PM
Security Audit — agent-trust-hub — remix-imessage-ad-from-sample