Murmell

About Murmell

Murmell is a shared cloud canvas that runs in the browser. A team and multiple AI coding agents work on the same repository, on a cloud machine. When you close your laptop, the agents keep going and the work eventually lands back in git.

Review

Murmell addresses a specific friction point: coordinating people and AI agents on a single codebase without overwriting each other's work. The tool combines a persistent cloud environment with file-level claims so multiple actors can edit concurrently. It supports a handful of agents today, with more planned.

Key Features

• Shared cloud canvas - A browser-based workspace where the team and agents operate on one repo, with a live preview.
• File-level claims - An agent claims a file before writing, preventing collisions when people and other agents are editing simultaneously.
• Multi-agent support - Run Claude Code, Codex, Kimi, and OpenCode together in the same room. Hermes and OpenClaw are on the roadmap.
• Persistent cloud machine - Work continues after you close the laptop. The environment stays alive, and agents don't depend on a local session.
• Git integration - All changes land back in git, so the project remains accessible through standard version control.

Pricing and Value

At launch, Murmell is free for the first week and includes $60 of Claude Code credits. Longer-term pricing is not yet defined. The current offer gives a short trial window, but teams evaluating it for ongoing use will need to wait for the pricing model to be announced.

Pros

• Agents keep working when no one is at the keyboard, so progress isn't tied to a developer's laptop uptime.
• File-level claims reduce the coordination overhead of deciding who edits what.
• Multiple agent types can run side by side, which lets a team compare outputs from different models in one session.
• The canvas includes a preview, making it possible to see front-end changes without leaving the tool.
• Git-native output means the work doesn't get siloed in a proprietary format.

Cons

• The agent roster is limited to four options today; teams that rely on other coding agents will have to wait for future additions.
• Mobile simulators (iOS, Android) are planned but not yet available, so mobile development previews still require external tools.
• Murmell is not well suited for teams that need agents to run inside a client's own environment or on-premise infrastructure-the current design is a cloud canvas, not a self-hosted or client-controlled machine.

Murmell fits teams that already use multiple AI coding agents and want a shared, persistent workspace where agents and people can commit to the same repo without manual file locking. It's less relevant for solo developers or teams that work exclusively within a single IDE and don't need an always-on cloud machine. The tool's usefulness will grow as the agent list expands and the mobile simulators ship.