computer-use-plus
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| CUP_MOCK | No | Set to '1' to run with a mock driver for testing without Windows UIA. | |
| COMPUTER_USE_PLUS_ROOT | No | Root directory of the project when not in the default location, used by DeepSeek Harness adapter. | |
| COMPUTER_USE_PLUS_AI_MODEL | No | Model name for the AI provider (e.g., gpt-4o-mini). Optional. | |
| COMPUTER_USE_PLUS_DATA_DIR | No | Directory for local memory, isolated agent, and runtime logs. Defaults to .data under the project directory. | .data |
| COMPUTER_USE_PLUS_AI_BASE_URL | No | Base URL for the AI provider API (e.g., https://api.openai.com/v1). Optional. | |
| COMPUTER_USE_PLUS_AI_KEY_FILE | No | Path to a file containing the AI provider API key. Optional; if not configured, local functionality still works. | |
| COMPUTER_USE_PLUS_EXECUTION_MODE | No | Execution mode. Windows default is backgroundOnly; set to foregroundAllowed for debugging old foreground mode. | backgroundOnly |
| COMPUTER_USE_PLUS_WECHAT_COMMAND | No | Command to launch WeChat for benchmark suite. Required only when using --execute. | |
| COMPUTER_USE_PLUS_RISK_POLICY_FILE | No | Path to a JSON policy file for high-risk action confirmations. | |
| COMPUTER_USE_PLUS_MINECRAFT_COMMAND | No | Command to launch Minecraft for benchmark suite. Required only when using --execute. | |
| COMPUTER_USE_PLUS_BROWSER_EXECUTABLE | No | Path to a browser executable for real benchmark suite execution. Required only when using --execute. |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| computer.stateB | 获取窗口、焦点和运行能力；includeUi=true 时一次返回可执行 UI 节点快照和短期 ref。 |
| computer.inspectB | 按需查询窗口或语义 UI 元素；默认不返回完整 UI 树。 |
| computer.waitB | 等待窗口或 UI 元素出现/消失，支持按标题、进程、角色、文本和 automationId 轮询。 |
| computer.screenshotA | 按需获取专用桌面窗口截图；默认只返回坐标元数据，避免增加 token。 |
| computer.invokeB | 优先入口：执行已保存 shortcut 或一组受限动作；高风险操作返回一次性确认令牌。 |
| shortcut.runC | 通过稳定 ID/名称运行已验证 shortcut，不向模型返回完整动作数组。 |
| computer.verifyA | 验证窗口、元素、CDP URL 或允许目录中的文件；返回每项断言的 expected、actual、passed。 |
| computer.cancelC | 取消一个或全部尚未确认的高风险操作。 |
| computer.actC | 在指定窗口中执行并验证一组声明式动作。 |
| computer.fastA | 可选的快速操作 AI：读取一次紧凑 UI 快照并批量执行当前动作，不写入长期记忆。使用与整理 AI 相同的可选 API key。 |
| computer.shortcutC | 由主 AI 显式保存、运行和整理动作链；本地脚本负责作用域隔离、聚类和清理。整理 AI 仅在 useAi=true 时调用。 |
| computer.executionC | 管理不影响用户前台桌面的 Windows 专用执行桌面。 |
| computer.browserB | 通过独立 profile 和 CDP/DOM Accessibility 后端操作 Chromium 页面；不会复用用户浏览器 profile。 |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
Latest Blog Posts
- Who's Calling? MCP Hosts Are an Identity Blind Spot (And the Spec Knows It)By Om-Shree-0709 on .mcpAgent IdentityOAuth 2.1
- Your AI Chatbot Just Exposed Your CEO's Salary to an InternBy Om-Shree-0709 on .Agent IdentityMCP SecurityOAuth Delegation
- Why MCP Servers Need Execution Sandboxing (And Why Your Current Stack Isn't Enough)By Om-Shree-0709 on .Agentic AiPrompt InjectionWebAssembly
MCP directory API
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/Ethanout/computer-use-plus'
If you have feedback or need assistance with the MCP directory API, please join our Discord server