JFrog Fly Desktop App
Overview
The JFrog Fly Desktop App is the central point for setting up and managing your local environment with JFrog Fly’s full capabilities. It bridges your local development tools with the JFrog Fly Registry and enables AI-powered workflows.
Platform Support
Currently supported: macOS
Windows and Linux support coming soon.
Installation
If you haven’t installed the Desktop App yet, see:
Desktop App Connection Guide →
Authentication
Login: Launch the app and sign in with your JFrog Fly credentials (GitHub, Google, or Email).
Logout: Select Sign Out from the app menu to disconnect.
Key App Functions
Connecting Your Package Managers to JFrog Fly
The app automatically detects and configures your local package managers (npm, pip, docker, maven, go, nuget, and more), connecting them to the Fly Registry so you can use standard commands without manual configuration.
Enable Agentic Workflows
The app includes support for JFrog Fly MCP (Model Context Protocol) - an industry-standard protocol that enables AI assistants to interact with development tools and services. With JFrog Fly MCP, you can use natural language commands in your IDE to manage packages, search releases, configure workflows, and deploy to runtime environments.
Supported Agentic IDEs
The Desktop App automatically integrates with supported AI-powered IDEs:
Currently Supported
- Cursor - Full JFrog Fly MCP integration
- VS Code Copilot - Full JFrog Fly MCP integration
Coming Soon
- Claude CLI
- WindSurf
Once the Desktop App detects a supported IDE, JFrog Fly MCP is automatically configured with no additional setup required.
What You Can Do with JFrog Fly MCP
JFrog Fly MCP enables natural language commands in your IDE for:
Package Management
- “Publish this npm package to Fly”
- “Find all Docker images in my registry”
- “Search for packages uploaded last week”
Release Management
- “Find the release where I fixed the login bug”
- “Show me details for release v2.3.1”
- “What did Jon deploy yesterday?”
Workflow Configuration
- “Configure this workflow to use Fly Registry”
- “Is my build workflow configured for Fly?”
- “Configure Fly for all workflows in this repo”
Deployment
- “Deploy the latest release to production”
- “Find the payment-service release and deploy to staging”
- “Show me the Docker image for this release”
Repository & Artifact Information
- “Show me all connected repositories”
- “What artifacts are in the backend repository?”
- “Get details for the frontend repo”
JFrog Fly MCP understands context and provides intelligent responses specific to your team, repositories, and artifacts.
Next Steps
Now that you understand the Desktop App, learn how to use it in your workflow:
- Local Artifact Management → - Upload and download packages locally
- Workflow Configuration → - Configure GitHub Actions workflows
- Runtime Deployment → - Deploy releases to Kubernetes