skill
oatda-skills
Claude Code plugin: 9 skills for OATDA unified LLM API gateway - text, vision, image, video, audio (TTS/STT/translation). 10+ providers, 1 API key.
Set up
Skills are installed by placing their folder where your agent loads skills.
The download comes from github.com/devcsde/oatda-skills itself, not from our server, so what you get is exactly what its author published.
Install instructions come from the README
README
synced Oct 27, 2026OATDA Skills Plugin for Claude Code
OATDA (One API To Direct All) provides unified access to 10+ LLM providers through a single API. This Claude Code plugin adds skills for text generation, vision analysis, image generation, video generation, and audio generation/transcription/translation.
Also available: OpenClaw skills · MCP server (Cursor, Cline, Claude Code) · /llms.txt for browsing agents
Supported Providers
| Provider | Chat | Image | Video | Audio |
|---|---|---|---|---|
| OpenAI | ✅ | ✅ (DALL-E, GPT-Image) | ✅ (Sora) | ✅ (TTS, Whisper) |
| Anthropic | ✅ | - | - | - |
| ✅ | ✅ (Imagen) | ✅ (Veo) | - | |
| Deepseek | ✅ | - | - | - |
| Mistral | ✅ | - | - | - |
| xAI | ✅ | ✅ | - | - |
| Alibaba | ✅ | ✅ (Qwen) | ✅ (Wan) | - |
| MiniMax | ✅ | ✅ | ✅ (T2V-01) | - |
| ZAI | ✅ | - | ✅ | ✅ (transcription) |
| Moonshot | ✅ | - | - | - |
Installation
Option 1: Load from local directory (development)
git clone https://github.com/devcsde/oatda-skills.git
claude --plugin-dir ./oatda-skills
Option 2: Install from marketplace
# Inside Claude Code:
1. add marketplace
/plugin marketplace add devcsde/oatda-skills
2. install skill
/plugin
then go to installed marketplaces, browse to and click "Install" on the OATDA Skills plugin
3. restart claude-code
then restart claude-code to load the skill/plugin
Option 3: Install via skills CLI (Claude Code, Cursor, Codex, and more)
npx skills add devcsde/oatda-skills
Configuration
Set your OATDA API key using one of these methods:
Environment variable (quickest)
export OATDA_API_KEY=your_api_key_here
Config file (recommended)
Run the setup script:
bash scripts/setup-credentials.sh
Or manually create ~/.oatda/credentials.json:
{
"version": 1,
"defaultProfile": "default",
"profiles": {
"default": {
"name": "default",
"apiKey": "your_api_key_here",
"createdAt": 1234567890,
"lastUsed": 1234567890
}
}
}
Important: Set secure permissions:
chmod 600 ~/.oatda/credentials.json
Available Skills
| Skill | Invocation | Description |
|---|---|---|
| Text Completion | /oatda:oatda-text-completion | Generate text using any of 10+ LLM providers |
| Vision Analysis | /oatda:oatda-vision-analysis | Analyze images using vision-capable models |
| Image Generation | /oatda:oatda-generate-image | Generate images from text descriptions |
| Video Generation | /oatda:oatda-generate-video | Generate videos (async, returns task ID) |
| Video Status | /oatda:oatda-video-status | Check video generation task status |
| Speech Generation | /oatda:oatda-generate-speech | Generate speech/audio from text (TTS) |
| Audio Transcription | /oatda:oatda-transcribe-audio | Transcribe audio recordings to text |
| Audio Translation | /oatda:oatda-translate-audio | Translate foreign-language audio to English text |
| Video Ad Creator | /oatda:oatda-video-ad-creator | Create branded video ads with your own brand, screenshots and voiceover (TTS + AI video + Remotion) |
| List Models | /oatda:oatda-list-models | List available models with filtering |
| Compare Models | /oatda:oatda-compare-models | Run one prompt through multiple LLMs in parallel and compare outputs side-by-side (max 8 models) |
| Check Balance | /oatda:oatda-check-balance | Check current balance, total usage, and remaining credits before expensive calls |
Claude will also invoke these skills automatically when relevant to your conversation.
Usage Examples
Text Generation
Write a haiku about code using claude
/oatda:oatda-text-completion Explain quantum computing using gpt-4o
Vision Analysis
Analyze this image: https://example.com/photo.jpg
Image Generation
/oatda:oatda-generate-image A cyberpunk city at night, neon lights, rain
Video Generation
/oatda:oatda-generate-video Ocean waves crashing on a beach at sunset
Then check status:
/oatda:oatda-video-status <task-id>
Speech Generation
/oatda:oatda-generate-speech Convert this announcement to speech using openai/gpt-4o-mini-tts with alloy voice
# Prefer: list models first — /oatda:oatda-list-models type=audio — then pick a current TTS id
Audio Transcription
/oatda:oatda-transcribe-audio Transcribe this meeting recording using openai/whisper-1
Audio Translation
/oatda:oatda-translate-audio Translate this French audio to English using openai/whisper-1
List Models
/oatda:oatda-list-models
What OpenAI models are available through OATDA?
Compare Models
/oatda:oatda-compare-models Compare how GPT-5, Claude Sonnet 4.5, and Gemini 3 Pro explain recursion
Check Balance
/oatda:oatda-check-balance Do I have enough credits for a 10-second video?
Plugin Structure
oatda-skills/
├── .claude-plugin/
│ └── plugin.json # Plugin manifest
├── skills/
│ ├── oatda-text-completion/
│ │ └── SKILL.md # Text generation skill
│ ├── oatda-vision-analysis/
│ │ └── SKILL.md # Image analysis skill
│ ├── oatda-generate-image/
│ │ └── SKILL.md # Image generation skill
│ ├── oatda-generate-video/
│ │ └── SKILL.md # Video generation skill
│ ├── oatda-video-status/
│ │ └── SKILL.md # Video status checking skill
│ ├── oatda-generate-speech/
│ │ └── SKILL.md # Text-to-speech audio generation skill
│ ├── oatda-transcribe-audio/
│ │ └── SKILL.md # Speech-to-text transcription skill
│ ├── oatda-translate-audio/
│ │ └── SKILL.md # Audio translation to English skill
│ ├── oatda-list-models/
│ │ └── SKILL.md # Model listing skill
│ ├── oatda-compare-models/
│ │ └── SKILL.md # Side-by-side LLM comparison skill
│ └── oatda-check-balance/
│ └── SKILL.md # Account balance / usage check skill
├── scripts/
│ └── setup-credentials.sh # API key setup helper
├── README.md
└── LICENSE
API Reference
All skills communicate with the OATDA REST API:
| Endpoint | Method | Purpose |
|---|---|---|
/api/v1/llm | POST | Text/chat completion (prompt) and vision analysis (contents[]) |
/api/v1/llm/image | POST | Deprecated alias for /api/v1/llm vision analysis (same body) |
/api/v1/llm/generate-image | POST | Image generation |
/api/v1/llm/generate-video?async=true | POST | Video generation |
/api/v1/llm/video-status/{taskId} | GET | Video status |
/api/v1/llm/speech | POST | Text-to-speech audio generation |
/api/v1/llm/transcriptions | POST | Speech-to-text transcription |
/api/v1/llm/translations | POST | Audio translation to English |
/api/v1/models | GET | List models |
/api/v1/llm/models?type=audio | GET | List audio-capable models |
Authentication: Authorization: Bearer <api_key> header on all requests.
Getting an API Key
- Visit oatda.com
- Sign up or log in
- Navigate to API Keys
- Click Generate New Key
- Store it securely (see Configuration above)
Security
- API keys are never logged or included in skill output
- Config file requires
chmod 600permissions - Only HTTPS image URLs are accepted (no HTTP, no local files)
- Internal/private IPs are rejected (localhost, 169.254.x.x, etc.)