ChatPPT

ChatPPT is an MCP server that offers AI-powered presentation generation through multiple access methods, including Python, Node.js, and a Streamable HTTP interface. The Streamable HTTP option lets clients interact with ChatPPT without installing dependencies by sending and receiving data over a lightweight, chunked HTTP protocol designed for MCP services. You can prototype quickly in clients that support MCP Streamable HTTP (e.g., Cursor, Trae) and progressively integrate richer capabilities such as PPT generation from themes, document uploads for automatic slide creation, and online editing before downloading the final output. If you prefer code-level integration, you can also run a local Node.js or Python client against the server to explore the full API surface and example calls.

Key capabilities include: generating PowerPoint presentations from given prompts or themes, uploading and processing documents (Word, PDF) to produce slide decks, and enabling online editing with a final download step. The server aims to provide a flexible workflow for creating, refining, and exporting presentations powered by AI models, with streaming HTTP support to minimize setup friction for clients that want to try the service immediately.

Prerequisites:

• Access to a compatible MCP environment or runtime (Node.js for Node-based usage, or Python environment if you plan to test via Python).
• Basic familiarity with Streamable MCP clients if you want to use the HTTP interface without dependencies.

Installation steps (example for Node.js runtime):

1. Ensure Node.js is installed on your machine (https://nodejs.org/).
2. Clone the ChatPPT MCP repository: git clone <repository-url> cd chatppt-mcp
3. Install dependencies (if applicable in your setup): npm install
4. Run the MCP server using the provided command configuration (adjust the path to your server script as necessary): node path/to/chatppt_server.js
5. If you are using the Streamable HTTP interface, ensure your client is configured to communicate with the server’s Streamable endpoint (the exact URL/port will be provided in your deployment context).

For Python workflows, adjust accordingly to run the Python-based client/server entry point if you have one in your copy of the repository. If you are only testing via Streamable HTTP, you can skip installing heavy dependencies and interact with the API directly.

Tips and caveats:

• Streamable HTTP is recommended for quick experimentation since it minimizes local environment setup and dependencies.
• If you see connection issues, verify that STREAMABLE_HTTP_ENABLED (or equivalent) is set in the environment and that the server is reachable at the configured endpoint.
• When using document uploads, ensure you are sending the correct content types and that size limits (if any) match your deployment constraints.
• For production deployments, consider configuring environment variables for API base URLs, authentication tokens, and any rate limits or access controls required by your organization.
• If you need to debug, run the server with verbose logging and inspect the MCP stream messages for request/response bodies and status codes.