Mannequin Context Protocols (MCPs) are rapidly turning into the usual for connecting AI brokers to exterior instruments, however the developer expertise hasn’t caught up. Discovery is fragmented, setup is clunky, and safety is just too usually bolted on final. Fixing this expertise isn’t a solo mission—it would take an industry-wide effort. A safe, scalable, and trusted MCP ecosystem calls for collaboration throughout platforms and distributors.
That’s why we’re excited to announce Docker MCP Catalog and Toolkit at the moment are accessible in Beta. The Docker MCP Catalog, now part of Docker Hub, is your start line for discovery, surfacing a curated set of well-liked, containerized MCP servers to jumpstart agentic AI improvement. However discovery alone isn’t sufficient. That’s the place the MCP Toolkit is available in. It simplifies set up, manages credentials, enforces entry management, and secures the runtime surroundings. Collectively, Docker MCP Catalog and MCP Toolkit give builders and groups an entire basis for working with MCP instruments, making them simpler to seek out, safer to make use of, and able to scale throughout tasks and groups.
We’re partnering with a few of the most trusted names in cloud, developer tooling, and AI, together with Stripe, Elastic, Heroku, Pulumi, Grafana Labs, Kong Inc., Neo4j, New Relic, Proceed.dev, and plenty of extra, to form a safe ecosystem for MCP instruments. With a one-click connection proper from Docker Desktop to main MCP purchasers like Gordon (Docker AI Agent), Claude, Cursor, VSCode, Windsurf, proceed.dev, and Goose, constructing highly effective, clever AI brokers has by no means been simpler.
This aligns completely with our mission. Docker pioneered the container revolution, remodeling how builders construct and deploy software program. As we speak, over 20 million registered builders depend on Docker to construct, share, and run fashionable purposes. Now, we’re bringing that very same trusted expertise to the following frontier: Agentic AI with MCP instruments.
Mannequin Context Protocol is gaining momentum — what enhancements are nonetheless wanted?
As MCPs grow to be the spine of agentic AI methods, the developer expertise nonetheless faces key challenges. Listed here are a few of the main hurdles:
Discovering the precise, official, and/or reliable instruments is difficult
Discovering MCP servers is fragmented. Builders search throughout registries, community-curated lists, and weblog posts—but it’s nonetheless arduous to know which of them are official and reliable.
Complicated installations and distribution
Getting began with MCP instruments stays advanced. Builders usually need to clone repositories, wrangle conflicting dependencies in environments like Node.js or Python, and self-host native providers—lots of which aren’t containerized, making setup and portability even more durable. On high of that, connecting MCP purchasers provides extra friction, with each requiring customized configuration that slows down onboarding and adoption.
Auth and permissions fall quick
Many MCP instruments run with full entry to the host, launched through npx or uvx, with no isolation or sandboxing. Credentials are generally handed as plaintext surroundings variables, exposing delicate knowledge and growing the danger of leaks. Furthermore, these instruments usually aren’t designed for scale and safety. They’re lacking enterprise-ready options like coverage enforcement, audit logs, and standardized safety.
How Docker might help remedy these challenges
The Docker MCP Catalog and Toolkit are designed to handle the above ache factors by securely streamlining the invention, set up, and authentication of MCP servers — making it straightforward to attach together with your favourite MCP purchasers.
Uncover and run MCP servers simply in safe, remoted containers
The MCP Catalog makes it straightforward to find and entry 100+ MCP servers — together with Stripe, Elastic, Neo4j, and plenty of extra — all accessible on Docker Hub. With the MCP Toolkit Docker Desktop extension, you’ll be able to rapidly and securely run and work together with these servers. By packaging MCP servers as containers, builders can sidestep widespread challenges reminiscent of runtime setup, dependency conflicts, and surroundings inconsistencies — simply run the container, and it really works.
Determine 1: Uncover curated and well-liked MCP servers in Docker MCP Catalog, a part of the Docker Hub
We’re not simply simplifying discovery and set up — we’re inserting safety on the coronary heart of the MCP expertise. As a result of MCPs run inside Docker container photographs, they inherit the identical built-in security measures builders already belief and a wealthy ecosystem of instruments for securing software program all through the availability chain. And we’re going additional. The Docker MCP Toolkit addresses rising threats distinctive to MCP servers like Software Poisoning and Software Rug Pulls, by leveraging Docker’s sturdy place as each a supplier of safe content material and safe runtimes.
Determine 2: The MCP Toolkit Docker Desktop Extension means that you can simply and securely run MCP servers in containers.
Go to the extensions menu of Docker Desktop to get began with Docker MCP Catalog and Toolkit, or use this for set up. Try our doc for extra info.
One-Click on MCP Consumer Integration with Constructed-In Safe Authentication
Whereas a curated checklist of MCPs and simplified safety is a superb start line, it’s only the start. You may join well-liked MCP servers from the Docker MCP Catalog to any MCP consumer. For purchasers like Gordon (Docker AI Agent), Claude, Cursor, VSCode, Windsurf, proceed.dev, and Goose, one-click setup will make integration seamless.
The Docker MCP Toolkit consists of built-in OAuth assist and safe credential storage, enabling purchasers to authenticate with MCP servers and third-party providers with out hardcoding secrets and techniques into surroundings variables. This ensures your MCP instruments run securely and reliably proper from the beginning.
Determine 3: Simply hook up with your favourite MCP purchasers like Gordon, Claude, Cursor, and proceed.dev with one click on.
Enterprise-Prepared MCP Tooling: Construct, handle, and share in Docker Hub
Quickly, you’ll be capable of construct and share your individual MCPs on Docker Hub—house to over 14 million photographs, hundreds of thousands of lively customers, and a sturdy ecosystem of trusted content material. Groups depend on Docker Hub for verified photographs, deep picture evaluation, lifecycle administration, and enterprise-grade tooling. Those self same trusted capabilities will quickly prolong to MCPs, giving groups entry to the newest instruments and a safe, dependable option to distribute their very own. And similar to container photographs, MCPs will combine with enterprise options like Registry Entry Administration and Picture Entry Administration, guaranteeing safe, streamlined developer workflows from finish to finish.
Wrapping up
Docker MCP Catalog and Toolkit deliver much-needed construction, safety, and ease to the fast-growing world of MCP instruments. By standardizing how MCP servers are found, put in, and secured, we’re eradicating friction for builders constructing smarter, extra succesful AI-powered purposes and brokers.
Whether or not you’re connecting to exterior instruments, customizing workflows, or scaling automation inside your IDE, Docker makes all the course of straightforward and safe. And that is only the start. With ongoing investments in increasing the MCP ecosystem and streamlining how instruments are managed, we’re dedicated to creating highly effective AI tooling accessible to each staff.
With Docker Catalog and Toolkit, your AI agent isn’t restricted by what’s in-built — it’s empowered by the whole lot you’ll be able to plug in.
Go to the extensions menu of Docker Desktop to get began with Docker MCP Catalog and Toolkit, or use this for set up. See it in motion throughout our upcoming webinar. Interested by internet hosting your MCP servers on Docker? Let’s join.
