Deskbrid
Let coding agents drive a Linux desktop—windows, input, screenshots, clipboard, print, and AT-SPI—with a large MCP tool surface across major DEs.
Overview
Deskbrid is a MCP server for the Build phase that provides 92+ Linux desktop automation tools including windows, input, screenshots, clipboard, print, and AT-SPI across nine DEs.
What is this MCP server?
- 92+ MCP tools for Linux desktop automation
- Window management, input, screenshots, clipboard, and print
- AT-SPI accessibility integration for UI targeting
- Support called out for nine desktop environments
- OCI container image ghcr.io/coe0718/deskbrid:0.12.3 with stdio transport in package metadata
- 92+ MCP tools advertised
- Nine desktop environments (DEs) supported per description
- OCI image ghcr.io/coe0718/deskbrid version 0.12.3
What problem does it solve?
Linux-based builders cannot reliably make coding agents operate real desktop apps without a sanctioned automation bridge.
Who is it for?
Indie devs on Linux who need agent-driven GUI automation, accessibility-aware UI control, or local RPA during product work.
Skip if: macOS or Windows-only shops, or teams unwilling to run high-privilege desktop control on developer machines.
What do I get? / Deliverables
Agents can invoke a broad Linux desktop tool suite for screenshots, input, and window control instead of fragile shell-only hacks.
- Agent-driven window and input control on Linux
- Screenshots, clipboard, and print actions via MCP
- Accessibility-tree driven UI automation through AT-SPI tools
Recommended MCP Servers
Journey fit
Desktop control integrations are built when you connect agents to real environments, which maps to Build integrations. 92+ OS-level tools are integration work: agents need sanctioned desktop APIs rather than manual clicking during dev.
How it compares
Linux desktop automation MCP (92+ tools), not a cloud deployer or LLM routing server.
Common Questions / FAQ
Who is Deskbrid for?
Linux developers and agent builders who need MCP tools to control windows, input, screenshots, clipboard, printing, and AT-SPI across common desktop environments.
When should I use Deskbrid?
Use it when integrating agents with native Linux GUIs for QA, demos, accessibility-driven navigation, or repeatable desktop workflows.
How do I add Deskbrid to my agent?
Pull ghcr.io/coe0718/deskbrid:0.12.3 (OCI) or follow the repo install path, configure stdio MCP in your client, and ensure the container or process can access your display/session per upstream docs.