Railgrid documentation
Build applications, run AI agents, and work with connected infrastructure through one shared platform.
AI prompt
Install the CLI, then install the Railgrid skill for your assistant.
railgrid skills installOne command writes the skill to ~/.claude/skills and ~/.agents/skills for Claude Code and Codex. Add --scope project for this project only, or --dir PATH for another client.
Then copy this prompt into your coding assistant.
Help me get started with the Railgrid SaaS hub. Ask which workspace I should use and what I want to build. If my organization self-hosts Railgrid, use its hub instead. Check whether the Railgrid CLI is installed. Guide me through login and workspace selection, then help me connect the workspace’s MCP server and inspect its available tools. Suggest a first task using the providers available to me.
MCP
Already logged in? Select your workspace, then print its MCP setup snippets.
kubectl railgrid use
kubectl railgrid mcp url --mcpserver-name defaultUse the printed configuration in your MCP client. It includes credentials; keep it private and review the server’s scope.
CLI
Install the CLI, then replace YOUR-HUB with your hub’s address.
kubectl railgrid --help
kubectl railgrid login --hub-url https://YOUR-HUB
kubectl railgrid useLogin updates your kubeconfig. Use the authentication guide if your hub requires a static token.
What will you build first?
Build and publish an application
Create a project, check its preview, and prepare a production deployment.
Start with App StudioCreate an AI agent
Choose a model, give it instructions, and try your first conversation.
Start with AI agentsConnect a cluster or server
Install an edge agent and connect through your hub.
Start with EdgesExplore what you can build with Railgrid
Each provider adds a capability to your workspace. Choose one to find its overview, quickstart, and guides.
App Studio
Build and publish applications.
Explore App StudioAI agents
Run assistants with tools, memory, and schedules.
Explore AI agentsEdges
Connect clusters, servers, and services.
Explore EdgesInfrastructure
Provision and manage template-based resources.
Explore InfrastructureCode
Connect GitHub and manage repositories.
Explore CodeKuery
Search your fleet and inspect resource relationships.
Explore KueryDatabricks
Connect tables and use their data.
Explore DatabricksBring the pieces of your work together
An application needs more than code. It needs a repository, a place to run, access to data, and people who can manage it. Railgrid brings those capabilities into a workspace so you can work across them from the console, CLI, or an AI assistant.
Start with the capability you need today. Add others as your workflow grows: connect a GitHub repository, provision a development environment, give an agent tools, or query a connected cluster.
Follow a connected workflow
Build an application using Databricks data shows how the parts work together. Use App Studio to build the application, Code for its repository, Infrastructure for its runtime, and Databricks for access to an existing table. The guide covers the required connections and grants as well as the application work.
How Railgrid fits your environment
A hub provides the shared control plane. Workspaces organize resources and access. Providers add capabilities such as App Studio, AI agents, and Edges. Your deployment determines which providers and external services are available.
Start with the Railgrid SaaS hub and choose a tutorial above; each guide lists its workspace prerequisites. To run Railgrid on your own infrastructure, see Self-hosting . Learn more about how Railgrid works .