Introducing JFrog’s MCP Server: Better vibes and easier AI automation
Making it easier to perform actions in the JFrog Platform using natural language
Good news! You no longer have to be a DevOps or JFrog expert to harness the power of the JFrog Software Supply Chain Platform. With the introduction of JFrog’s MCP Server, we’re making the JFrog Platform accessible to your favorite large language models (LLMs).
Now, every developer can take advantage of the detailed security and package information available in JFrog, such as vulnerability data from the JFrog Catalog, without needing to context-switch. This increased awareness, thanks to MCP, makes everyday work easier for everyone.
What is MCP?
If you haven’t been keeping up with the latest in AI, MCP (short for Model Context Protocol) is an open standard integration framework designed to connect AI systems with external tools, data, and services using a consistent, lightweight interface. It eliminates the need for building one-off connectors by providing a unified way to expose internal tool capabilities to large language models.
The beauty of MCP is its interoperability. Any MCP client can connect to any MCP server, as shown in the diagram below:
MCP General Architecture
What’s included in the new JFrog MCP Server?
The JFrog Platform delivers DevOps, Security, MLOps, and IoT services across your software supply chain. Our new MCP Server enhances its accessibility, making it even easier to integrate into your workflows and the daily work of developers. Here’s a high-level overview of what’s included today (with more to come over the next few months):
Tools
In the beginning phases of this roll-out, we’re starting with a core set of essential tools that provide users with the ability to:
- Create and view projects, repositories, and more
- Get detailed information from JFrog, including the vulnerability status of open-source packages
- Understand what’s in use inside your organization
These are just the beginning. We’ll be rolling out many additional tools over time, so stay tuned!
Clients
The JFrog MCP Server typically works with an MCP client that supports OAuth, such as VSCode, Cursor, and Claude. This means you can bring any AI agents adhering to MCP and interact seamlessly with JFrog.
Remote
One of the most significant advantages of JFrog’s MCP Server is its “remote” delivery. This means there’s nothing for you to install, and it’s always up to date with the latest tools. Plus, it’s available for every JFrog SaaS customer, with no cloud restrictions.
Let’s bring the automating power of JFrog’s MCP to life with an example
Here’s an example of how you might use JFrog’s MCP to create a new project named green-pizza that supports Docker and Maven.
Without MCP: Creating a new project like green-pizza in JFrog manually means going through multiple steps. First, you have to create the project itself. Then, for each package type, Docker and Maven, you need to set up three repositories: local, remote, and virtual. That’s six repositories in total, each requiring configuration. After that, you need to assign them to the project, set permissions, and validate the setup. It’s powerful, but slow and prone to human error.
With MCP: Just say, “Hey, create a new project named green-pizza that supports Docker and Maven.” And that’s it. The AI instantly spins up the project, creates and configures all necessary repositories, and connects everything properly, no manual steps and no guesswork required. It’s DevOps without the friction.
The bottom line is this: with MCP, you can just say what you need and the AI client together with MCP will understand and perform all the steps needed to get you there.
How to get started
It’s easy to get started. Simply visit the JFrog MCP Server documentation page and follow the instructions provided.
You can also check out this demo video to learn how easy it is to use JFrog’s MCP server to enhance and accelerate your JFrog Platform workflows with generative AI. We show you the seamless set-up process and how you can get started triggering actions within the platform leveraging your favorite LLMs.
We’re excited to announce that the JFrog Remote MCP Server is now in open beta for all JFrog SaaS customers. Ready to simplify your software development process and integrate AI into your JFrog workflows? Start your free trial or enable the beta of JFrog’s MCP Server now!