Skills
skills/pedronauck/skills/promo-video/Gen Agent Trust Hub

promo-video

Fail

Audited by Gen Agent Trust Hub on Mar 18, 2026

Risk Level: HIGHEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it ingests untrusted data from the local repository (git logs, README, and source code) to determine video content without sanitization or boundary markers. * Ingestion points: SKILL.md (Phase 1) reads git history and project files. * Boundary markers: Absent. * Capability inventory: Extensive, including npm, npx, ffmpeg, and python execution. * Sanitization: Absent.
  • [EXTERNAL_DOWNLOADS]: The skill downloads assets and code from external sources. It fetches audio from the ElevenLabs API, downloads royalty-free music from Bensound and Pixabay using curl, and bootstraps projects using npx create-video@latest. These target well-known services.
  • [COMMAND_EXECUTION]: The skill executes multiple shell commands for media processing and project management. This includes ffmpeg and ffprobe for audio/video handling, npx remotion studio for previews, and whisper for transcription tasks.
Recommendations
  • HIGH: Downloads and executes remote code from: unknown (check file) - DO NOT USE without thorough review
Audit Metadata
Risk Level
HIGH
Analyzed
Mar 18, 2026, 01:53 AM