proliferate
The open-source AI IDE for Claude Code, Codex, OpenCode, and more. Run agents in parallel, locally or in the cloud, and build reusable workflows.
What is it?
What it is
Open-source AI IDE that runs coding agents (Claude, Codex, OpenCode, Cursor, Grok) in parallel, each with isolated git worktrees, terminals, and conversation state.
Why it exists
Enables parallel AI agent workflows with isolated workspaces, shared integrations (MCPs, skills, tools), and event-driven recurring agent runs across a self-hostable control plane.
Who should use it
Developers and teams looking to run multiple AI coding agents in parallel with isolated worktrees and integrated tools.
Who should avoid it
Users who need a simple IDE without AI agent orchestration or those who cannot comply with AGPL-3.0 licensing.
How it works
A quick walkthrough in plain English
How proliferate works
Step 1 of 3
You interact with it
Open proliferate, send a request, or connect it to your stack.
Features
Advantages
- Multi-agent parallelism increases development throughput by running multiple AI agents simultaneously
- Isolated worktree architecture prevents branch conflicts and state pollution between tasks
- Supports a wide variety of AI coding agents from different providers
- Flexible deployment options from self-hosted to major cloud platforms
- Shared integrations and tools reduce configuration duplication across agents
- Workflow scheduling enables automated repetitive tasks like code reviews and triage
- Community Discord and documentation available for support
- TypeScript codebase with clear contribution guidelines
Disadvantages
- AGPL-3.0 license imposes strict copyleft requirements for modifications and distributions
- Desktop-focused application with limited web-based access
- Development requires multiple toolchains: Rust stable, Node.js 22+, pnpm, Python 3.12+, Docker
- Full-stack local development is complex with multiple services and profiles
- Relatively early-stage project with modest star count (441) suggesting limited community adoption
- macOS primary platform may have limited Windows/Linux support
- Self-hosting requires significant infrastructure and maintenance overhead
Installation
cloud
AWS one-click: self-hosted-aws.md is a CloudFormation wrapper that provisions the stack on EC2
native
Run from source: make install, make dev-local. Full-stack: make server-install, make setup PROFILE=main, make build, make dev-list, make run PROFILE=main. Requirements: Rust stable, Node.js 22+, pnpm, Python 3.12+, uv, Docker
compose
Docker Compose: self-hosted-deploy.md runs Caddy, Postgres, and the API, with bootstrap and update scripts
FAQ
What is Proliferate and what does it do?
Proliferate is an open‑source AI IDE that lets you run multiple coding agents (Claude Code, Codex, OpenCode, Cursor, Grok, and more) in parallel within a single workspace. Each task gets its own isolated git worktree, terminal, conversation, and review state, so agents don’t interfere with one another.
How are multiple agents managed and isolated?
Proliferate uses Git worktree workspaces – each task runs in its own isolated branch and working directory. This provides separate git histories, terminals, and conversation states for every agent, preventing cross‑talk and keeping work cleanly organized.
What AI agents are supported and how does Proliferate run them?
It supports Claude, Codex, OpenCode, Cursor, Grok, and others via native harnesses. Each agent runs through its own harness, enabling native features like Computer Use, Browser Use, MCPs, skills, and custom tools that are shared across all agents.
How can I self‑host Proliferate?
The full Proliferate control plane is self‑hostable. You can deploy it via Docker Compose, AWS (one‑click CloudFormation), GCP, Azure, Kubernetes, or air‑gapped setups. The deployment docs cover all options, and you configure the desktop app to point at your instance using the provided guides.
How do I run Proliferate from source for development?
Clone the repo, install Rust stable, Node.js 22+, pnpm, and Python 3.12+ with uv (and Docker for the local control plane). Run `make install` and `make dev-local` to start the desktop app with the bundled AnyHarness runtime. For a full local stack, run `make server-install`, `make setup PROFILE=main`, then `make build` and `make run PROFILE=main`.
Featured in Videos
YouTube tutorials and walkthroughs for proliferate
Alternatives
Similar projects ranked by category, topics, and text overlap.
