varg-dropship-reel

Pass

Audited by Gen Agent Trust Hub on Nov 1, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill downloads the 'Bangers-Regular.ttf' font file from Google's official fonts repository on GitHub to support caption rendering.
  • [COMMAND_EXECUTION]: The skill invokes several local binaries to process media, including ffmpeg and ffprobe for video assembly, and node/bun for running auxiliary scripts. It also uses playwright to automate browser screenshots of product pages.
  • [INDIRECT_PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection as it ingests and processes untrusted data from external product links and supplier websites.
  • Ingestion points: scripts/screenshot.mjs accepts arbitrary URLs (product or store links) provided by the user or extracted from supplier data.
  • Boundary markers: No specific delimiters or "ignore instructions" headers are used when processing external product descriptions or site content.
  • Capability inventory: The skill possesses capabilities for network access (curl to api.varg.ai), file system operations (reading/writing media clips), and command execution (ffmpeg, playwright).
  • Sanitization: While the skill uses standard media processing libraries, it lacks explicit sanitization to prevent the AI from following instructions that might be hidden in the text or metadata of the processed websites.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 1, 2026, 12:01 PM