blnk

blnk

A powerful terminal-based chat application for seamless interaction with multiple AI providers.

Features

• Multi-provider support (OpenAI, Anthropic, Google Gemini)
• Rich CLI interface with markdown formatting
• Dynamic model switching and API selection
• MCP integration for advanced reasoning
• Streaming responses with progress indicators
• Command system with help documentation
• YAML configuration with automatic JSON conversion
• Secure API key storage in user's home directory

Installation

Requires Python 3.13 or higher.

1. Clone the repository:

git clone https://github.com/frgmt0/blnk.git
cd blnk

2. Run the installation script:

./install.sh

This will:

• Create necessary config directories
• Install dependencies
• Set up the default configuration
• Install blnk in development mode

Quick Start

1. Run the setup command to configure your API providers:

blnk
/setup

2. Follow the prompts to configure your providers and API keys.

Usage

Start blnk:

blnk

Available Commands

• /help - Show help message
• /apis - List available AI providers
• /use <api> - Switch to specified AI provider
• /model-list - Show available models for current/all providers
• /switch-model - Switch to a different model for current provider
• /tools - List available MCP tools
• /exit - Exit the application

Examples

/use anthropic                               # Switch to Anthropic provider
/switch-model claude-3-5-sonnet-20241022    # Switch to Claude 3 model
/tools                                      # Show available tools

Configuration

blnk stores its configuration in ~/.blnk/config/:

• config.yaml - Main configuration file (edit this for changes)
• config.json - Runtime configuration (automatically generated)
• mcp/*.json - MCP server configurations

MCP Integration

blnk supports integration with MCP (Model Context Protocol) servers. While native MCP-reasoner support is planned for future releases, you can already configure and use MCP servers:

1. Create a JSON config file in ~/.blnk/config/mcp/ (e.g., default.json)
2. Configure the server details:

{
    "server": {
        "command": "node",
        "args": ["path/to/mcp-reasoner/dist/index.js"],
        "env": {}
    }
}

Multiple MCP servers can be configured by creating additional JSON files in the mcp/ directory.

The configuration includes:

• Default API provider
• Model preferences
• API keys (securely stored)
• Token limits
• MCP settings

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

1. Fork the repository
2. Create your feature branch (git checkout -b feature/AmazingFeature)
3. Commit your changes (git commit -m 'Add some AmazingFeature')
4. Push to the branch (git push origin feature/AmazingFeature)
5. Open a Pull Request

License

This project is licensed under the MIT License - see the LICENSE file for details.

Acknowledgments

• Built with MCP-Reasoner for advanced reasoning capabilities
• Thanks to all contributors

Author

frgmt0 - [email protected]