Skip to main content
skill CX Codex

gbro-collage-broll — Claude Code skill for collage video

The gbro-collage-broll Claude Code skill turns a short voiceover or abstract idea into a vertical paper-collage video. It pauses for approval of the visual concept and still frame before rendering.

MIT 100 installs ★ 61 updated 2026-10-24

Install into Claude Code · Skills CLI (Vercel) All methods and platforms ↓
npx skills add akillness/jeo-skills --skill gbro-collage-broll -a claude-code
Open on GitHub

What gbro-collage-broll does

The skill maps a short spoken line or abstract concept to an editorial collage animation. Its workflow has three approval stops: the visual metaphor, a still frame, and the rendered video. It waits for confirmation at each stop.

Run bash scripts/check_setup.sh to validate the environment. The setup needs Python >= 3.10, google-genai >= 2.10.0, GEMINI_API_KEY, ffmpeg, and ffprobe; the Python package belongs in ~/hyperframes-projects/.omni-venv. Gate 2 also relies on Codex’s built-in image_gen. Missing any of these components blocks the corresponding check or generation step.

Project files include visual-spec.json, contact sheets, and QA notes. The video is generated by scripts/generate_video.py, which uses gemini-omni-flash-preview by default; the resulting MP4 is converted to a silent format. The skill does not handle exact frame-by-frame layer edits, transparent assets, or prompt-only requests without a rendered clip.

What it helps with

  • turn a short voiceover into a visual metaphor
  • create a vertical video from paper-collage imagery
  • review a still frame before video rendering
  • save contact sheets and QA notes for generated frames

Example prompts for the agent

  • Turn this short voiceover into collage B-roll; propose the visual concept first.
  • Make a vertical paper-collage video from this abstract idea.
  • Prepare a still for the collage clip, then stop for my approval.

How to install gbro-collage-broll

1Skills CLI (Vercel)

npx skills add akillness/jeo-skills --skill gbro-collage-broll -a claude-code

2Manually: put the files in place

~/.claude/skills/gbro-collage-broll/SKILL.md

Copy the whole skill folder; for one project use .claude/skills/gbro-collage-broll/

Step by step, with commands for each platform: How to install skills in Claude Code: five ways

Questions about gbro-collage-broll

How do I install gbro-collage-broll in Claude Code?

Run npx skills add akillness/jeo-skills --skill gbro-collage-broll -a claude-code. The skill file is placed at ~/.claude/skills/gbro-collage-broll/SKILL.md.

What does gbro-collage-broll do?

It turns a brief voiceover or abstract concept into a vertical paper-collage animation. Video rendering follows approval of the visual metaphor and still frame.

What does gbro-collage-broll need to run?

It requires Python >= 3.10, google-genai >= 2.10.0, GEMINI_API_KEY, ffmpeg, and ffprobe. Gate 2 also requires Codex’s built-in image_gen.

Source text

SKILL.md · MIT
---
name: gbro-collage-broll
description: >
  Turn a 5-second voiceover script, opinion statement, or abstract concept into an editorial halftone paper-collage assemble-from-empty B-roll video. Use when the user requests "collage b-roll", "paper collage B-roll", "halftone collage", "collage video", "gbro-collage-broll", or asks to convert voiceover lines into animated editorial collage visual metaphors. Enforces a 3-gate approval workflow (metaphor confirmation -> still frame confirmation -> video generation via Gemini Omni Flash).
allowed-tools: Bash Read Write Edit Glob Grep
permissions: [env, network]
compatibility: Runs in Codex environment (Gate 2 depends on built-in image_gen). Requires Python >= 3.10, google-genai >= 2.10.0, GEMINI_API_KEY, and ffmpeg/ffprobe.
metadata:
  tags: video-generation, b-roll, paper-collage, halftone, gemini-omni-flash, editorial-animation
  version: "1.0"
---

# gbro Collage B-roll

把一句约 5 秒的口播压成一个 sharp visual idea,再做成高级编辑风纸拼贴组装动画。

## When to use this skill

Use this skill whenever:
- User asks for "collage b-roll", "paper collage B-roll", "halftone collage", "纸拼贴 b-roll", "半调拼贴", or "gbro-collage-broll".
- Converting voiceover lines or abstract concepts into editorial halftone paper-collage visual metaphors.
- Generating assemble-from-empty 9:16 vertical video clips via Gemini Omni Flash with strict 3-gate human approval.

When NOT to use this skill:
- User requires frame-by-frame exact layer manipulation, timeline editing, or transparent assets (use HyperFrames / HTML-based video renderers instead).
- User only wants video prompts without rendering final MP4 clips.
- User requests realistic human or product photoreal commercial footage.

## Instructions

### 1. Prerequisites & Environment Check

Before Gate 1, run the setup validator script:

```bash
bash scripts/check_setup.sh
```

Ensure `GEMINI_API_KEY`, `ffmpeg`/`ffprobe`, Python >= 3.10, and `google-genai >= 2.10.0` (in `~/hyperframes-projects/.omni-venv`) are ready.

### 2. 3-Gate Approval Workflow

#### Gate 1: Visual Metaphor Approval
- Compress the ~5s script into one clear visual thesis (core meaning, emotion, action verb, 3–6 key objects, background/accent colors, assembly sequence).
- Output the metaphor proposal without generating images/videos, then STOP and await user approval.

#### Gate 2: Still Frame Approval
- After Gate 1 approval, write `<project>/visual-spec.json` and generate the 9:16 still frame via Codex `image_gen`.
- Generate contact sheet `<project>/still-contact-sheet.jpg` and write `<project>/gate2-qa.md`. STOP and await user approval before calling any video model.

#### Gate 3: Video Generation & QA
- After Gate 2 approval, prepare empty color first frame (`<project>/first-frame.png`) and target last frame (`<project>/last-frame.png`).
- Execute batch Gemini Omni Flash video generation using `scripts/generate_video.py` (defaulting to `gemini-omni-flash-preview`):

```bash
~/hyperframes-projects/.omni-venv/bin/python \
  scripts/generate_video.py \
  --batch <project>/omni-jobs.json \
  --concurrency 3
```

- Convert output MP4 to silent format (`<project>/final-5s-noaudio.mp4`), run frame extraction contact sheets, and write `<project>/gate3-qa.md`.

## Examples

### Example 1: Single Line B-roll Request
Input:
> `collage b-roll: 很多人以为 AI 是来替你思考的,其实它更像一面镜子,会把你问题里的漏洞照出来。`

Workflow:
1. Gate 1 proposal:
   - 核心意思: AI 是照出漏洞的镜子而非替代思考
   - 隐喻: 放大镜与撕纸镜面透出光芒照亮排版漏字
   - 色彩: 芥末黄底,黑白半调,深红点色
2. Gate 2 generation & contact sheet review.
3. Gate 3 Omni Flash video generation & silent MP4 delivery.

## Best practices

1. **Strict 3-Gate Discipline**: Never skip Gate 1 or Gate 2. Stopping for human approval saves API cost and aligns aesthetic direction before running video models.
2. **Color Palette Diversity**: Do not default all clips to cobalt blue. Use orange, mustard yellow, dark purple, teal, or forest green based on semantics.
3. **Assemble-from-Empty Timing**: Ensure Image 1 is an empty color field and Image 2 is the exact completed frame so Omni Flash performs crisp stop-motion assemble animation.
4. **No Text/UI Clutter**: Keep still frames clean from readable letters, logos, watermarks, or UI artifacts.

## References

- Project repo: [https://github.com/pyang5166/gbro-collage-broll](https://github.com/pyang5166/gbro-collage-broll)
- Default Video Model: `gemini-omni-flash-preview`

© akillness · MIT · akillness/jeo-skills

Similar skills

All skills for Design →

From the same set99

The whole set →
okf okf writes and validates Google Open Knowledge Format v0.1 bundles — a folder of Markdown atoms with YAML frontmatter any agent can read. CX Codex +1 127installs agentic-skills agentic-skills runs a task through six gated phases, from spec to ship, and keeps code from starting before the requirements are frozen. CX Codex +1 125installs x-twitter-scraper The x-twitter-scraper skill accesses X data through Xquik for post search, user lookups, media retrieval, monitoring, and approval-based account actions. CX Codex +1 124installs deep-research deep-research turns a topic into an extendable outline, sends parallel agents through each item, then writes a markdown report. Built for comparing a set of things across shared fields. CX Codex +1 120installs webtoon-harness The webtoon-harness skill takes a webtoon episode from trend research through panel creation to a vertical-scroll viewer. Use it to coordinate a full episode workflow. CX Codex +1 115installs opencut A skill for Claude Code guides work on the OpenCut video editor repository. It activates for local setup, Rust/WASM core questions, and contribution planning. CX Codex +1 114installs devup-ui Sets up Devup UI zero-runtime CSS-in-JS in React applications, connects build plugins, and guides migrations away from runtime styling engines. CX Codex +1 111installs paperbanana paperbanana turns method text or a paper into diagrams, statistical plots and figure reviews, picking the cheapest mode that still gets the job done. CX Codex +1 108installs

Project sponsors

Companies that keep this analytics open