AI Media Generator
Platform-correct prompts for AI image, video and music, plus browser automation
Test report
- Verdict
- Tested · Works
- Score
- Tested
- Jul 16, 2026
- Environment
- Claude Code 2.x (agent harness)
- Upstream re-checked
- Aug 10, 2026 · 69e08c2
⚠ The author changed this skill after we tested it. The verdict below describes the version we ran on Jul 16, 2026; it's queued for a re-test.
Task: write a prompt for a 10-second pour-over coffee commercial. Baseline (from memory, without skill) produced a typical 'cinematic 10-second commercial... beautiful, slow-motion, dynamic camera movement, 4K, ultra detailed, masterpiece' — measured by grep: 4 junk words from the rubric's stop-list, 0 named camera movements, 0 optics, 0 negative prompts; on the skill's own 10-point scale, this was 3 raw points minus 4 penalties = 0/10, i.e., below its own submission threshold of '≥8'. With the skill, it was 10/10: 0 junk words, specific subject (matte terracotta funnel, 60° conical ribs), 4 scene elements, 100mm f/4 + dolly-in + rack focus, negatives from negative-bank.md addressing real artifacts (deformed product, morphing shape, inconsistent label). The main gain was structural: selector.md shifted from single-stage t2v to two-stage i2v (first hero shot in Flux, then animation in Kling) — precisely the technique the repo calls the root cure for product deformation, while baseline went straight into this pitfall. Installation from a bare HOME works (SKILL.md is in the repo root, clone directly into ~/.claude/skills), all relative links in SKILL.md resolve — 0 broken ones out of those checked. Caveat: did not generate the media itself (paid accounts needed), so evaluated the prompt artifact, not the final video, and the rubric is the skill's own rubric.
Scored on four weighted criteria — install, triggering, output vs. baseline, docs. How scoring works
- Installs cleanly 4/5
- Triggers reliably 5/5
- Output vs. baseline 8/10
- Docs & honesty 4/5
What AI Media Generator does
Picks the right generation platform (Kling, Veo, Suno, Midjourney, Flux, Seedream, Runway, Nano Banana) and builds the prompt in a six-layer structure with a ten-point pre-send scoring gate, then optionally drives the site via browser automation. Triggers on AI image/video/music requests, storyboards, MV work, character consistency, or fixing flawed generations.
How to install AI Media Generator
mkdir -p ~/.claude/skills
git clone https://github.com/Hao0321/ai-media-generator.git ~/.claude/skills/ai-media-generator
Skills live in ~/.claude/skills/ (global) or .claude/skills/
(per-project). Restart Claude Code after installing.
Commands — how to trigger AI Media Generator
-
/ai-media-generatorPlatform-correct prompts for AI image, video and music, plus browser automation
It also activates on plain-language prompts like these:
-
Write me a Kling prompt for a 10 second pour over coffee commercial -
I need a Suno prompt structured properly for a music video concept -
Help me fix this AI video generation that keeps deforming the product shape
Frequently asked questions
- Is the AI Media Generator skill free?
- Yes. The skill itself is free from Hao0321/ai-media-generator. SkillProof publishes the install command and an independent test verdict at no cost.
- Does AI Media Generator work with Claude Code?
- We tested it with Claude Code 2.x (agent harness) on Jul 16, 2026. Verdict: Tested · Works. Task: write a prompt for a 10-second pour-over coffee commercial. Baseline (from memory, without skill) produced a typical 'cinematic 10-second commercial... beautiful, slow-motion, dynamic camera movement, 4K, ultra detailed, masterpiece' — measured by grep: 4 junk words from the rubric's stop-list, 0 named camera movements, 0 optics, 0 negative prompts; on the skill's own 10-point scale, this was 3 raw points minus 4 penalties = 0/10, i.e., below its own submission threshold of '≥8'. With the skill, it was 10/10: 0 junk words, specific subject (matte terracotta funnel, 60° conical ribs), 4 scene elements, 100mm f/4 + dolly-in + rack focus, negatives from negative-bank.md addressing real artifacts (deformed product, morphing shape, inconsistent label). The main gain was structural: selector.md shifted from single-stage t2v to two-stage i2v (first hero shot in Flux, then animation in Kling) — precisely the technique the repo calls the root cure for product deformation, while baseline went straight into this pitfall. Installation from a bare HOME works (SKILL.md is in the repo root, clone directly into ~/.claude/skills), all relative links in SKILL.md resolve — 0 broken ones out of those checked. Caveat: did not generate the media itself (paid accounts needed), so evaluated the prompt artifact, not the final video, and the rubric is the skill's own rubric.
- What is the AI Media Generator SkillProof Score?
- 8.4/10 — installs cleanly 4/5, triggers reliably 5/5, output vs. baseline 8/10, docs & honesty 4/5.
- How do I install AI Media Generator?
- Copy the install command from this page, run it in your terminal, and restart Claude Code. Skills live in ~/.claude/skills/ (global) or .claude/skills/ inside a project.
- Can I use AI Media Generator with Cursor, Copilot, Gemini CLI, Codex or other AI tools?
- The SKILL.md format is native to Claude (Claude Code, Desktop, claude.ai). The instructions inside adapt to other assistants: Cursor rules, GitHub Copilot instructions, Windsurf rules, Custom GPTs, AGENTS.md for OpenAI Codex, and GEMINI.md for Google Gemini CLI — our conversion guides cover each, and the free converter on the tools page does the wrapping for you.