daan

Dáàn is a lightweight client designed to interact with OpenAI-compatible models and external tools via MCP (Model Context Protocol). This MCP integration enables Daan to harness external knowledge and utilities when generating responses, and it supports running miniapps or small MCP services that can be invoked within conversations. With Daan, you can configure different model backends, switch between OpenAI-compatible APIs, and manage personalized chat personas, all while leveraging MCP to extend capabilities through tool execution and data access. To use MCP-enabled features, obtain an API key, configure your chosen model/provider, and start the Daan server so it can expose MCP endpoints for tooling and context enrichment. You can then create or enable MCP services (miniapps) to perform tasks or fetch information during chats, which Daan can call as part of the conversation flow.

Prerequisites:

• Node.js (recommended latest LTS) and npm installed on your system
• Optional: Git for cloning the repository

Installation steps:

1. Clone the repository: git clone https://github.com/your-organization/daan.git cd daan

2. Install dependencies: npm install

3. Configure API keys and environment variables as needed. For example, provide your OpenAI API key in the environment or a .env file if supported by the project.

4. Start the server: npm run build || npm run start

5. Verify the server is running on the expected port (default may be http://localhost:3000 or as configured).

If you prefer running directly with Node (without a build step): node path/to/server.js

Note: Replace paths and commands with the actual entry points used by the project if different from the examples above.

Tips and considerations:

• Ensure you have a valid OpenAI API key or equivalent provider credentials for MCP model access.
• When configuring MCP services (miniapps), you may need to publish or reference tooling in a supported format; consult the project docs for how to register and invoke these services.
• Keep your API keys secure; avoid committing them to version control and consider using environment variables or secret managers.
• If you encounter token usage concerns, use Daan’s token estimation features to monitor consumption per conversation.
• If MCP services require network access, ensure your runtime environment allows outbound calls to the configured APIs.
• Review any licensing notes (e.g., AGPL-3.0) to ensure compliant use in your environment, especially for commercial deployments.