openchamber/openchamber

Project overview

Its Session Goals can keep an agent working after the app is closed until a goal completes, and it can run the same task in up to five model sessions and fuse the strongest parts into a new session, all inside one workspace spanning desktop, web/PWA, VS Code, iOS, Android, and CLI/server surfaces.

Project type

AI Agent · AI Coding · Workflow

Use cases

Coding & Development · Automation

Deployment

Refer to project documentation

License

MIT

Best for

• Developers and AI engineers who need to run, monitor, review, and ship AI coding work across multiple devices.
• Teams that work through GitHub issues and pull requests and want agents to receive failed checks and review comments.
• Users who want to track session states, approvals, token use, and costs across projects.

Key capabilities

• Runs AI coding sessions, shows what changed, and supports taking the work through review and release.
• Keeps an agent working toward a defined goal until completion, blocking, or a configured limit.
• Runs the same task in up to five separate model sessions and can combine the strongest parts into a new session.
• Turns a large diff into an AI-guided, ordered tour that groups related edits and explains how they fit together.
• Lets users inspect a running app beside the conversation and send an element screenshot, styles, position, and browser errors to the agent.
• Starts sessions from GitHub issues or pull requests, sends failed checks and review comments to the agent, and updates or merges pull requests.
• Makes projects and sessions available across desktop, web/PWA, VS Code, iOS, and Android, including progress review and terminal reattachment.
• Connects paired devices through Private Relay with one-time QR codes, encrypted connections, and revocation support.

Limitations and risks

• CLI/Web and VS Code use an installed OpenCode CLI; the desktop bundle avoids a separate installation by including it.
• The documentation advises using LAN mode only on a trusted network and protecting browser access with a UI password.

Getting started

• Download the latest desktop release; on Linux, make the AppImage executable and run it (Linux may require FUSE). The desktop bundle includes a matching copy of the OpenCode CLI, so no separate installation is needed.
• Install the CLI/Web package, which requires Node.js 22+, run the install command, and set a UI password, then open the CLI/Web interface. This path uses an installed OpenCode CLI.

Evidence and sources

• GitHub project description: Agentic Development Environment based on OpenCode AI agent
• README: **OpenChamber is an open-source workspace for running and reviewing AI coding work on desktop, web, VS Code, and mobile.**
• README: Give a session a finish line with **Session Goals**. OpenChamber checks the result after every turn and keeps the agent working until it completes the goal, gets blocked, or reach…
• README: Use **Multi-run** to give the same task to up to five models, each in its own session and optionally its own worktree. See what each one actually built, choose the best result, or…
• README: **Changes Walkthrough** turns a large diff into an AI-guided tour of the change. It groups related edits into steps, puts them in the order the change makes sense, and explains ho…

Source updated at: 2026-09-22T18:07:09Z

Open original GitHub project