💡 Summary
A CLI tool that critiques motion/design videos and provides actionable feedback using AI.
🎯 Target Audience
🤖 AI Roast: “Powerful, but the setup might scare off the impatient.”
Risk: Medium. Review: shell/CLI command execution; outbound network access (SSRF, data egress); API keys/tokens handling and storage; filesystem read/write scope and path traversal; dependency pinning and supply-chain risk. Run with least privilege and audit before enabling in production.
video-agent-skills
A CLI tool that critiques motion/design videos using Gemini, plus a Claude Code skill bundle for one-command installs.
What it does
- Uploads a video to Gemini and returns blunt, production-focused feedback
- Surfaces strengths, issues, and concrete improvements
- Works as a CLI and as a Claude Code Skill
Install (CLI)
./scripts/install.sh
Then run:
video-agent-skills feedback -video /path/to/video.mp4
Control harshness with -tone:
video-agent-skills feedback -video /path/to/video.mp4 -tone nice video-agent-skills feedback -video /path/to/video.mp4 -tone super-harsh
Reverse engineer the likely prompt used to create a finished video:
video-agent-skills reverse -video /path/to/video.mp4
Configure API key
The tool reads GEMINI_API_KEY or GOOGLE_API_KEY from your environment. You can also place it in:
~/.config/video-agent-skills/.env
Example .env:
GEMINI_API_KEY=your_key_here
CLI usage
video-agent-skills feedback -video <path> [options]
video-agent-skills reverse -video <path> [options]
Options:
-video string Path to a video file (required)
-model string Gemini model name (default: gemini-3-flash-preview)
-prompt string Prompt override (feedback or reverse)
-tone string Feedback harshness: nice|normal|harsh|super-harsh (default: harsh)
-api-key string Gemini API key (overrides GEMINI_API_KEY/GOOGLE_API_KEY)
Claude Code Skill (local install)
./scripts/install-claude-skill.sh
This installs to:
~/.claude/skills/video-agent-skills
Claude Code Skill (npm install)
npm install @dunctk/video-agent-skills
The installer downloads a prebuilt binary for your platform. If you want to force a source build:
VIDEO_AGENT_SKILLS_FORCE_BUILD=1 npm install @dunctk/video-agent-skills
Supported prebuilt targets:
- darwin/amd64
- darwin/arm64
- linux/amd64
- linux/arm64
- windows/amd64
Example feedback

Pros
- Provides blunt, actionable feedback
- Supports multiple installation methods
- Allows customization of feedback tone
Cons
- CLI interface may be intimidating for beginners
- Requires API key configuration
- Feedback may be too harsh for some users
Related Skills
pytorch
S“It's the Swiss Army knife of deep learning, but good luck figuring out which of the 47 installation methods is the one that won't break your system.”
agno
S“It promises to be the Kubernetes for agents, but let's see if developers have the patience to learn yet another orchestration layer.”
nuxt-skills
S“It's essentially a well-organized cheat sheet that turns your AI assistant into a Nuxt framework parrot.”
Disclaimer: This content is sourced from GitHub open source projects for display and rating purposes only.
Copyright belongs to the original author dunctk.
