JFrog AI Catalog Overview
Welcome to the JFrog AI Catalog – your centralized hub for AI asset discovery, governance, and deployment. Here you can view and manage models, MCP servers, and related AI assets in your organization, including both those you are currently using, and those you want to use.
AI Challenges
How Can the AI Catalog Help You?
The JFrog AI Catalog simplifies the process of integrating AI into your projects. It is your centralized hub for discovering, governing, and securing all AI assets in your organization.
The AI Catalog bridges the gap between MLOps, Platform Engineering, and Development by providing a single source of truth for the two critical pillars of modern AI:
- AI Models: The intelligent engines themselves (for example, GPT-4, Llama 3, or custom internal models).
- Model Context Protocol (MCP) Servers (Tools): Connectors that give AI agents access to your systems so they can read files, query data, call APIs, and take other actions through governed tools.
- COMING SOON: Skills Registry: Skills are versioned, reusable bundles containing instructions, scripts, and assets that enable AI agents to execute repetitive tasks with centralized governance and security controls.
The JFrog MCP Registry is the system of record for MCP servers in the JFrog Platform. You allow servers per project, define tool policies, and scan packages before developers use them. Developers connect coding agents to approved MCP servers through the JFrog MCP Gateway (JFrog CLI plugin), which acts as a secure local proxy. For more information, see MCP Registry Overview.
Whether you are deploying a model to production or connecting a coding agent to internal systems through MCP, the AI Catalog helps ensure every asset is vetted, secure, and compliant. It enables you to find, govern, and deploy AI models quickly and securely, and to govern MCP servers with the same rigor, streamlining your workflow.
Shadow AI detection helps you find unmanaged models in use across your organization and manage them securely.
Whether you are new to JFrog or seeking to leverage advanced features, this guide will help you navigate core capabilities and essential functions, making your AI integration process straightforward and secure. You will gain insight into managing external models, model packages, and custom models, and into governing MCP servers, so you maintain control over what can run within your organization.
AI Catalog supports four main types of AI assets:
| AI Asset Types | Description |
|---|---|
| External Models | Accessed via third-party APIs (for example, OpenAI, and Google Gemini) |
| Model Packages | Utilized from public repositories (for example, Hugging Face) |
| Custom Models | Internally developed models tailored to your organization's needs |
| MCP Servers | Servers implementing the Model Context Protocol to provide tools to AI agents. |
These assets need governance, have many versions, and may carry risk. The AI Catalog allows you to manage them in a unified way.
AI Catalog Features and Benefits
See also:
- JFrog AI Catalog Architecture
- AI/ML Overview Dashboard
- Detect Shadow AI
- AI Catalog Quick Start
- Discover and Allow Models
- Connect AI Providers
- MCP Registry Overview
- Get Started with MCP
Updated 1 day ago