The MCP Dev Summit featured more than 50 sponsors offering MCP and related agentic AI products for the enterprise. Here are some offerings:
- StackLock provides a complete runtime, security, and governance platform for MCP servers, including an MCP gateway, registry, runtime guardrails, and a developer/admin portal. Developers can deploy their MCP servers using the StackLock runtime and control access to the servers using the MCP gateway, which includes authentication, auditing, and single sign on (SSO) capabilities
- pgEdge offers an Agentic AI Toolkit with an MCP server that works with most versions of Postgres. It gives LLMs and agents secure access to Postgres databases. DBAs can use it to chat with the DB, query the schema for rows, columns, and indexes and analyze usage.
- FastMCP is an open source Python framework to build MCP servers, and sponsors an open source platform called Prefect Horizon on which to deploy and run the MCP servers in production.
- Reboot develops and deploys MCP Apps as either front end only or front end/back end apps using React and Python (and soon Typescript. Reboot supports ChatGPT, Claude, VS Code, and Goose. MCP Apps are deployed as iframes in an LLM’s browser space.
User types in the App automatically register as an MCP endpoint, and user methods are automatically exposed as MCP tools. Methods on other types can be exposed manually. UI methods implement MCP tools with resources, for example a pointer to a React app. - AugmentCode analyzes application source code to produce a knowledge graph that provides semantic context about the application and exposes the context using an MCP server. Semantic context improves results for LLM based coding agents and IDEs.
- Docker’s MCP Catalog and Toolkit provides a set of curated MCP servers running in containers. The Toolkit handles setup, authentication, and enterprise grade security for the MCP servers.
The MCP Gateway acts as a centralized proxy between AI agents and multiple MCP servers.
