A unified interface and SDK for interacting with multiple AI providers across all modalities including text, image, audio, video, and embeddings.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install genai-calling
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install genai-calling using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is genai-calling?
The genai-calling skill provides a streamlined, provider-agnostic way to integrate high-level generative AI capabilities into your local or cloud-based workflows. By using this Openclaw Skills contribution, developers can bypass the complexity of managing individual API client libraries for OpenAI, Google, Anthropic, and more, opting instead for a single standardized command-line or programmatic interface.
Built with portability in mind, it supports robust environment-based configuration and multi-modal interactions. Whether you are generating images, transcribing audio with Whisper, or performing complex text analysis, genai-calling ensures your AI agent has a consistent and reliable bridge to the world's most powerful LLMs.
genai-calling Use Cases
- Rapid prototyping of multi-modal AI applications using a single unified SDK.
- Automating image generation and speech-to-text transcriptions via CLI scripts.
- Building custom AI agents that require standardized access to multiple LLM providers.
- Implementing local MCP (Model Context Protocol) servers for secure, authenticated AI workflows.
How genai-calling Works
- The user configures provider API keys (OpenAI, Anthropic, etc.) via environment variables or .env files.
- The genai-calling runtime initializes a client that resolves the specified model string format (provider:model_id).
- Input data, such as prompts or file paths for media, is processed and sent to the respective AI provider's endpoint.
- The skill manages the request lifecycle, including timeouts and multi-modal output handling.
- Results, including generated text or paths to created media files, are returned to the user or calling application.
genai-calling Setup
Install the runtime package using pip:
python -m pip install --upgrade genai-calling
Configure your environment by creating a .env.local file in your skill directory:
# Example for OpenAI
OPENAI_API_KEY=your_api_key_here
GENAI_CALLING_TIMEOUT_MS=120000
Test the installation with a simple text prompt:
uvx --from genai-calling genai --model openai:gpt-4o-mini --prompt "Hello"
genai-calling Data Schema & Taxonomy
The skill uses a structured approach for request and response handling, ensuring consistency across different modalities.
| Component | Format | Description |
|---|---|---|
| Model Identifier | {provider}:{model_id} |
Example: openai:gpt-4o-mini or openai:whisper-1. |
| Input Parts | Message Array | Supports text, image-path, audio-path, and video-path parts. |
| Output Spec | Modalities List | Defines expected return types (text, image, audio, etc.). |
| Configuration | ENV Variables | Managed via GENAI_CALLING_* prefix or provider-specific keys. |
genai-calling Advanced Features
- Multi-modal support including text-to-speech (TTS), speech-to-text (STT), and image generation.
- Built-in MCP Server (
genai-mcp-server) for standardized tool-use and agent communication. - Advanced authentication rules using
GENAI_CALLING_MCP_TOKEN_RULESfor granular access control. - Private URL protection with optional bypass for local-only binary data downloads.
- Hierarchical environment variable loading from local, production, and user-wide config files.
SKILL.md
Loading
Related Openclaw Skills
Frontend CV
A sophisticated resume engine that transforms raw career data into high-quality, print-ready HTML and PDF portfolios.
Skill Dedup Scanner
A diagnostic utility for scanning AI agent workspaces to detect similar or duplicate skills that may cause model confusion.
Video URL to Text (Toncall)
An automated pipeline to download video, extract audio, and generate transcriptions using Volcengine ASR for Openclaw Skills users.
Fontpreview
A fast, local-first documentation utility for Fontpreview patterns and developer best practices.
Md Convert
A local-first documentation reference tool providing instant access to technical guides, implementation patterns, and development best practices.
HK Stock To Gougoubi Prediction
A specialized tool that transforms Hong Kong stock analysis into structured, deterministic prediction market proposals for the Gougoubi platform.
Featured*
