Overview

Membrane makes your software talk to anything else. This documentation describes how.

Getting Started

Pick your use case:

Give Claude, ChatGPT, Cursor, or any agent the ability to connect to apps, run actions, and react to events.

Add integrations to your product so your tenants can connect their apps, sync data, and automate workflows.

Use the API, SDK, or CLI to build scripts, internal tools, and automations.

Membrane is fully programmable and API-first. Everything is accessible through multiple interfaces:

Give coding agents Membrane capabilities with one command.

Manage integrations, sync config, and launch MCP servers.

Call any Membrane endpoint directly.

TypeScript/JavaScript SDK with typed accessors.

Expose Membrane as tools for AI agents via MCP.

Pre-built React components and embeddable screens.

Web UI for managing your workspace and monitoring.

Membrane provides a set of building blocks for integrating with external apps:

How Membrane organizes elements, functions, and layers to connect your app with external software.

Authenticated links to external apps.

Run operations in external apps.

Subscribe to changes in external apps.

Logs, alerts, error inspection, and debugging tools for every integration touchpoint.

Deploy Membrane on your own infrastructure with Docker and Helm charts.

Move workspace configuration between environments.

Security practices, data privacy, and compliance.