superglue

superglue's MCP server exposes discoverability and execution capabilities for pre-built superglue tools. Through the MCP interface you can query available tools, inspect their inputs and outputs, and invoke them in a controlled, tool-driven workflow. The MCP does not allow ad-hoc tool creation or modifications; instead you interact with a catalog of pre-built tools designed to integrate with any API, database, or file storage system, with built-in handling for authentication, documentation, and data mapping. Use the MCP to orchestrate tool executions in production environments and to empower agents or internal GPTs to run established, self-healing tools across your enterprise data sources.

Once running, you can discover tools, inspect their schemas, and execute them with the provided inputs. The MCP is designed for agentic use cases and internal GPTs requiring reliable, pre-built tooling access. If you run in a self-hosted setup, you’ll typically interact with the MCP through its API and accompanying documentation to integrate with your existing systems and governance practices.

Prerequisites:

• Docker or a compatible container runtime installed on your host
• Basic familiarity with running containerized services

Install and run:

1. Pull and run the Docker image for the Superglue MCP server:

   docker run -d --name superglue-mcp -i superglueai/superglue

2. Verify the container is running and exposed as needed by your environment (check docker ps and logs):

   docker ps docker logs -f superglue-mcp

3. If you need to customize environment variables, pass them at run time. Example:

   docker run -d --name superglue-mcp -e API_KEY=your_api_key -p 8080:8080 -i superglueai/superglue

4. Follow the MCP guide for production installation details and endpoint configuration: https://docs.superglue.cloud/mcp/using-the-mcp

5. Optional: Integrate with your deployment pipeline or orchestrator according to your infrastructure (Kubernetes, etc.).

Notes and tips:

• The MCP exposes discovery and execution for pre-built superglue tools only; new tool-building or ad-hoc integration creation is not supported via MCP.
• When self-hosting, ensure your environment has appropriate network access and security controls around API endpoints and data flow.
• Refer to the MCP Guide for detailed API references, authentication requirements, and usage patterns: https://docs.superglue.cloud/mcp/using-the-mcp
• If you use the Docker image, you can leverage Docker Compose or Kubernetes for orchestrated deployment and scaling.
• Check for updates to the superglue MCP image regularly to benefit from tool improvements and security patches.