datadam_mcp

DataDam is a persistent memory layer that runs as an MCP server and backs its data with Supabase. It exposes a set of tools under the mcp/tool namespace for storing, retrieving, updating, and deleting personal data across categories like digital_products, basic_information, interests, and more. You can connect multiple AI tools to the same DataDam instance via HTTP stream or stdio, enabling consistent context retention across conversations and tools. The server is designed to work without authentication yet, so avoid storing sensitive information and plan to add OAuth later.

To use it, run the server and configure your client(s) to talk to the MCP endpoint. Your client can invoke tools such as datadam_search_personal_data, datadam_extract_personal_data, datadam_create_personal_data, datadam_update_personal_data, and datadam_delete_personal_data to manage records. If your workflow involves multiple AI tools, you can keep one personal database that all tools query through these MCP endpoints. The data model groups records into categories with optional tags, and queries can be filtered accordingly.

The tools are designed to be invoked from your AI tooling environment, either by HTTP requests over the streamable MCP transport or via stdio to a local process, depending on your deployment choice. The docs provide examples of how to search by title, extract a category, create a new record, or update an existing one, enabling a cohesive memory layer across your AI stack.

Prerequisites:

• Node.js installed on the host (recommended version per your project)
• Access to a Supabase project (URL and anon/public key)
• (Optional) Git for cloning the repository

1. Install dependencies (if this project is hosted on npm/yarn):

npm install
# or
yarn install

2. set up environment variables

• Create a .env file or export environment variables in your hosting environment:

export SUPABASE_URL="https://xyzcompany.supabase.co"
export SUPABASE_ANON_KEY="your-anon-key"
export PORT=3000

3. Start the MCP server

node path/to/server.js

4. Verify the server is running

• Check the console for a listening message on the configured port
• Test connectivity via HTTP stream or stdio as described in the docs

5. Optional: Run the server in the background or in a container

• Docker: run with appropriate environment variables (if you prefer containerized deployment)
• PM2 or systemd can manage the Node process for uptime guarantees

Tips and common issues:

• There is no auth yet; do not store highly sensitive data. Plan for OAuth when available.
• Ensure SUPABASE_URL and SUPABASE_ANON_KEY are correctly configured; invalid credentials will cause data access failures.
• If the server does not respond, verify that the PORT is not blocked by your hosting environment and that the Supabase backend is reachable.
• When migrating data or testing, use the datadam_create_personal_data and datadam_extract_personal_data tools to validate the data model and permissions.
• For local development, you can run the server via stdio (path/to/server.js) and connect clients accordingly; for multi-client usage, HTTP stream is recommended.
• Keep your dependencies updated and monitor the MCP server logs for any deprecation warnings or API changes.