---
slug: "happy-claude-skills"
source_type: "readme"
source_url: "https://cdn.jsdelivr.net/gh/iamzhihuix/happy-claude-skills@main/README.md"
repo: "https://github.com/iamzhihuix/happy-claude-skills"
source_file: "README.md"
branch: "main"
---
[English](https://github.com/iamzhihuix/happy-claude-skills/blob/HEAD/README.md) | [中文](https://github.com/iamzhihuix/happy-claude-skills/blob/HEAD/README.zh-CN.md)

# Happy Claude Skills

A collection of practical skill plugins for AI coding agents.

Works with **Claude Code**, **Codex**, **Factory Droid**, **OpenClaw**, **Cursor**, and [40+ agents](https://github.com/vercel-labs/skills#available-agents).

## Included Skills

### docx-format-replicator
Extract formatting from existing Word documents and generate new documents with the same format.

**Use Cases:**
- Corporate document template replication
- Batch generation of consistently formatted documents
- Technical specification document series
- Standardized documentation like development task sheets

### video-processor
Download and process videos from YouTube and other platforms. Supports video download, audio extraction, format conversion, and Whisper transcription.

**Use Cases:**
- Download videos from YouTube and other platforms
- Extract audio from video files
- Convert videos to MP4/WebM formats
- Transcribe audio/video to text using Whisper

### wechat-article-writer
Automated WeChat article writing workflow with 4 steps: research, writing, title generation, and formatting optimization.

**Use Cases:**
- Write WeChat official account articles
- Generate viral headlines
- Content creation for self-media
- Article formatting and optimization

### trends-bulletin
Multi-platform trending topics bulletin. Fetches trends from HuggingFace, GitHub, Hacker News, Product Hunt, Reddit, and YouTube, then sends a formatted report to Telegram.

**Use Cases:**
- Daily multi-platform trend monitoring
- AI/tech hot topics discovery
- Content creator topic research
- Telegram automated trend push

### browser
Browser automation using Chrome DevTools Protocol. Start Chrome, navigate pages, execute JavaScript, take screenshots, and interactively pick DOM elements.

**Use Cases:**
- Web scraping with authenticated sessions
- Visual regression testing
- DOM inspection and data extraction
- Screenshot capture for documentation

### 1password
Manage passwords and API credentials using 1Password CLI (`op`). Save, query, and retrieve API keys/tokens, inject secrets into scripts via environment variables.

**Use Cases:**
- Save API keys and tokens to 1Password vault
- Query and retrieve stored credentials
- Inject secrets into scripts with `op run --env-file`
- Manage API credentials with metadata (created date, expiry, tags)

### happy-image-gen
Universal AI image generation — one CLI across **8 providers**: OpenAI DALL-E / gpt-image, Google Gemini / Imagen, Replicate (Flux / SDXL / any), Stability AI, FAL, Ark Seedream, Bailian (qwen-image / wanx), SiliconFlow (Kolors).

**Use Cases:**
- Draft illustrations and hero images
- Text-on-image posters (Ark Seedream for Chinese, OpenAI gpt-image-1 for English)
- Reference-driven edits (OpenAI gpt-image / Google Gemini multimodal)
- Swap providers with a single `--provider` flag

### happy-video-gen
Universal AI video generation — one CLI across **10 providers**: OpenAI Sora, Google Veo, Runway, Pika, Luma, FAL (Kling / Wan / wrappers), Ark Seedance, Bailian Wanx, MiniMax Hailuo, Vidu. Handles the submit → poll → download pipeline uniformly.

**Use Cases:**
- Text-to-video or image-to-video clips (5–10s)
- Quickly A/B different providers for the same prompt
- Cheap drafts via FAL Kling Turbo or Ark Seedance Lite
- Voice-synced dialogue via Google Veo 3 or OpenAI Sora 2

### happy-audio-gen
Universal AI TTS / voice generation — one CLI across **6 providers**: OpenAI TTS, ElevenLabs, Bailian (qwen-tts / qwen3-tts-vd with auto-chunking for long Chinese text), MiniMax speech-02-hd, SiliconFlow CosyVoice, PlayHT 2.0.

**Use Cases:**
- Narration / voice-over for scripts
- Long-form Chinese article → audiobook (auto sentence-split)
- Multilingual TTS with voice cloning (ElevenLabs, PlayHT)
- Fast English preview reads (OpenAI gpt-4o-mini-tts)

### happy-dreamina
ByteDance Jimeng (Dreamina) image and video generation via the official `dreamina` CLI. Browser-based login (no API key needed), covers text2image / image2image / text2video / image2video plus task history via `list_task`.

**Use Cases:**
- Jimeng-exclusive models for image and short video
- Browser OAuth login instead of managing another API key
- Resuming async jobs via submit_id

### open-source-prep
Prepare a private project for open-source release. Scans for leaked secrets in code and git history, recommends a license based on project context, and generates all required open-source scaffolding.

**Use Cases:**
- Pre-push secret scan (GitHub / AWS / OpenAI / Anthropic / Slack / Stripe tokens, private keys, `.env` files)
- License selection via 3-question decision tree (defaults to MIT; Apache 2.0 for enterprise or patent-heavy projects)
- Auto-generate LICENSE, CONTRIBUTING.md, SECURITY.md, and README disclaimer
- Audit `.gitignore`, bundle identifiers, and `package.json` for trademark / ownership issues

## Installation

### Universal Install (Recommended)

Works with Claude Code, Codex, Droid, OpenClaw, Cursor, and 40+ agents:

```bash
# Install all skills (interactive — pick agents and skills)
npx skills add iamzhihuix/happy-claude-skills

# Install specific skills
npx skills add iamzhihuix/happy-claude-skills --skill happy-image-gen --skill happy-video-gen
npx skills add iamzhihuix/happy-claude-skills --skill browser --skill 1password

# Install to a specific agent
npx skills add iamzhihuix/happy-claude-skills --agent codex
npx skills add iamzhihuix/happy-claude-skills --agent droid
npx skills add iamzhihuix/happy-claude-skills --agent openclaw

# Install all skills to all agents, no prompts
npx skills add iamzhihuix/happy-claude-skills --all
```

### Claude Code Plugin Marketplace

```
/plugin marketplace add iamzhihuix/happy-claude-skills
/plugin install browser@happy-claude-skills
/plugin install happy-image-gen@happy-claude-skills
/plugin install happy-video-gen@happy-claude-skills
/plugin install happy-audio-gen@happy-claude-skills
/plugin install happy-dreamina@happy-claude-skills
```

### Local Development

```bash
git clone https://github.com/iamzhihuix/happy-claude-skills.git

# Claude Code
claude --plugin-dir /path/to/happy-claude-skills

# Or use skills CLI for any agent
npx skills add ./happy-claude-skills --all
```

## Usage

After installation, simply describe your needs — your agent will automatically identify and invoke the appropriate skill:

> "I have a document template and need to generate 5 new documents with the same format"

> "Download this YouTube video and transcribe it to text"

> "Help me write a WeChat article about AI programming tips"

> "Scrape the product information from this webpage"

> "帮我发一下热词速报" / "Send me a trends bulletin"

> "Save this API key to 1Password"

## Dependencies

### docx-format-replicator
- Python 3.7+
- python-docx

```bash
pip install python-docx
```

### video-processor
- Python 3.7+
- yt-dlp
- FFmpeg
- openai-whisper

```bash
pip install yt-dlp openai-whisper
brew install ffmpeg  # macOS
```

### trends-bulletin
- Bun（无外部依赖，使用内置 fetch）

```bash
# 安装 Bun（如未安装）
curl -fsSL https://bun.sh/install | bash
```

### browser
- Node.js 18+
- puppeteer-core
- Google Chrome

```bash
npm install --prefix skills/browser
```

### happy-image-gen / happy-video-gen / happy-audio-gen
- Bun 1.1+ (`curl -fsSL https://bun.sh/install | bash`, or run via `npx -y bun`)
- Per-provider API keys (exported as env vars or referenced via `1password` skill). See each skill's `references/providers.md` for the full list.

### happy-dreamina
- ByteDance `dreamina` CLI: `curl -fsSL https://jimeng.jianying.com/cli | bash`
- A Jimeng account (`dreamina login` opens the browser for OAuth — no API key)

## Project Structure

```
happy-claude-skills/
├── AGENTS.md                    # Agent instructions (Codex, etc.)
├── .claude-plugin/
│   └── marketplace.json         # Claude Code marketplace config
├── skills/
│   ├── docx-format-replicator/
│   │   ├── SKILL.md             # Skill definition
│   │   ├── scripts/             # Python scripts
│   │   ├── assets/              # Example files
│   │   └── references/          # Reference docs
│   ├── video-processor/
│   │   ├── SKILL.md             # Skill definition
│   │   └── scripts/             # Python scripts
│   ├── wechat-article-writer/
│   │   └── SKILL.md             # Skill definition
│   ├── trends-bulletin/
│   │   ├── SKILL.md             # Skill definition
│   │   └── scripts/             # Python script
│   ├── browser/
│   │   ├── SKILL.md             # Skill definition
│   │   ├── package.json         # Node.js dependencies
│   │   └── scripts/             # Node.js scripts
│   ├── happy-image-gen/         # 8-provider image CLI (Bun)
│   ├── happy-video-gen/         # 10-provider video CLI (Bun, async)
│   ├── happy-audio-gen/         # 6-provider TTS CLI (Bun)
│   └── happy-dreamina/          # Pure instruction skill wrapping dreamina CLI
├── README.md
└── LICENSE
```

## Acknowledgments

- **video-processor** skill is adapted from [claude-code-hooks-multi-agent-observability](https://github.com/disler/claude-code-hooks-multi-agent-observability) by [@disler](https://github.com/disler)
- **browser** skill is based on [Mario Zechner](https://mariozechner.at)'s article [What if you don't need MCP?](https://mariozechner.at/posts/2025-11-02-what-if-you-dont-need-mcp/) ([GitHub](https://github.com/badlogic/browser-tools)), adapted from [Factory.ai](https://docs.factory.ai/guides/skills/browser)

## Contributing

Issues and Pull Requests are welcome!

## License

MIT License