ai-skills
Banuba/ai-skills
Banuba Agent Skills for Claude Code, Codex & Qwen — install Face AR & Video Editor & Photo Editor SDK knowledge into your AI coding assistant in one CLI command.
Instalação
npx convez-ai Banuba/ai-skills1 comando no terminal: detecta o formato do repositório (marketplace ou plugin avulso) e instala via Claude Code. Requer Node 18+ e o Claude Code instalado.
Métricas
README
A tradução para o português deste README ainda está na fila.
Banuba Face AR, Video Editor and Photo Editor SDKs — Agent Skills
Banuba Agent Skills are portable knowledge packs that install into Claude Code, Codex, and Qwen Code, giving the AI assistant offline access to Banuba Face AR, Video Editor and Photo Editor SDKs documentation plus autonomous scaffolding for new Face AR Web, VE and PE SDKs projects.
Give your AI coding assistant expert-level knowledge of Face AR SDK, Banuba Video Editor and Photo Editor SDKs. Build mobile and web apps by describing what you want.
What Are Agent Skills?
Agent Skills are portable knowledge packs that plug into AI coding assistants. By installing the Banuba agent skills, you get:
- Documentation lookup skills are local-first and can answer from bundled docs. Builder skills may access the network to fetch versioned LLM docs, clone official samples, and install platform dependencies.
- Guided code generation — a step-by-step walkthrough and explanation of SDK implementation
- Autonomous scaffolding — a builder agent that creates complete Video Editor or Photo Editor projects from scratch
Why Agent Skills??
Dropping a raw llms.txt dump into your assistant works once — but it bloats every prompt with the entire SDK surface, goes stale the moment Banuba ships a new release, and leaves the assistant to guess at structure. Agent skills are loaded on demand via slash commands, so context stays focused on the file you're actually editing. They're versioned and pinned to a specific SDK release, so generated code matches the API you're integrating against. And unlike a passive doc dump, skills carry executable behavior — the build skills can apply starter-kit templates, wire up dependencies, and scaffold an entire project, not just answer questions about one.
Available Skills
| Skill | Description |
|---|---|
far-general | Build Face AR apps on Web, Android, iOS, and Desktop (C++), get explanations, read documents, and more. |
build-video-editor | Implement features, write code, and set up Banuba Video Editor SDK projects |
build-photo-editor | Implement features, write code, and set up Banuba Photo Editor SDK projects |
explain-video-editor-photo-editor-docs | Look up Banuba Video and Photo Editor SDKs docs — configuration, UI customization, guides |
The build skills include starter kit templates for common platforms that let you autonomously create complete Banuba Video Editor, Photo Editor, or Face AR applications — applying templates and implementing features end-to-end.
Supported Platforms
- Face AR SDK — Web, Android, iOS, Desktop C++ (Claude Code only)
- Video & Photo Editor SDKs — Android, iOS, Flutter, React Native
SDK Versions
- Face AR SDK:
- Web: v1.18.2
- Android: v1.18.2
- iOS: v1.18.2
- Desktop C++: v1.18.2
- Video Editor SDK
- Android: v1.52.0
- iOS: v1.52.1
- Flutter: v0.43.0
- React Native: v0.50.0
- Photo Editor SDK
- Android: v1.3.8
- iOS: v1.3.6
- Flutter: v0.5.0
- React Native: v0.4.0
Setup Instructions
Claude Code
Install using the Vercel Skills CLI:
# Install all skills
npx skills add Banuba/ai-skills -a claude-code
# Or install a specific skill only
npx skills add Banuba/ai-skills --skill build-video-editor -a claude-code
# List available skills first
npx skills add Banuba/ai-skills --list
Skills are installed into the .claude/skills/ directory and loaded automatically.
Codex
Install using the Vercel Skills CLI:
npx skills add Banuba/ai-skills -a codex
Skills are installed into the .codex/skills/ directory.
Qwen Code
Install using the Vercel Skills CLI:
npx skills add Banuba/ai-skills -a qwen-code
Skills are installed into the .qwen/skills/ directory.
Face AR SDK
Install using the Vercel Skills CLI:
# Install only the far-general skill
npx skills add Banuba/ai-skills --skill far-general -a claude-code
# See the available Banuba skills first
npx skills add Banuba/ai-skills --list
Usage
Once installed, invoke skills with slash commands in your AI coding assistant:
Build a feature
/far-general Build a Snapchat-style face-filter camera for Web, Android, iOS, or Desktop
/build-video-editor Set up a Video Editor for Android with AI Clipping
/build-photo-editor Add Photo Editor with AR filters to my iOS app
Look up documentation
/explain-video-editor-photo-editor-docs How do I customize the export settings?
Get explanation
/far-general How do I load a custom .zip effect at runtime in a Web project?
How It Works
Each documentation skill bundles the complete SDK guides and API references locally. Skills read directly from these bundled files — no external services or MCP servers are required.
The build skills include starter kit templates for common use cases like video editors or photo editors. They detect your project's platform and generate code accordingly.
Project Structure
.agents/skills/ Portable skill definitions (SKILL.md per skill)
.claude/skills/ Claude Code skills + bundled documentation
.codex/skills/ Codex skills + bundled documentation
.qwen/skills/ Qwen Code skills + bundled documentation
The .agents/ directory contains tool-agnostic skill definitions. The .claude/, .codex/, and .qwen/ directories each contain the same skills and documentation packaged for their respective AI coding assistant.