智能桌面代理 是 AI Skill Hub 本期精选MCP工具之一。综合评分 7.5 分,整体质量较高。我们推荐使用将其纳入你的 AI 工具库,帮助提升工作效率。
智能桌面代理 是一款遵循 MCP(Model Context Protocol)标准协议的 AI 工具扩展。通过 MCP 协议,它可以让 Claude、Cursor 等主流 AI 客户端直接访问和操作外部工具、数据源和服务,实现 AI 能力的无缝扩展。无论是文件操作、数据库查询还是 API 调用,都可以通过自然语言在 AI 对话中直接触发,极大提升生产效率。
智能桌面代理 是一款遵循 MCP(Model Context Protocol)标准协议的 AI 工具扩展。通过 MCP 协议,它可以让 Claude、Cursor 等主流 AI 客户端直接访问和操作外部工具、数据源和服务,实现 AI 能力的无缝扩展。无论是文件操作、数据库查询还是 API 调用,都可以通过自然语言在 AI 对话中直接触发,极大提升生产效率。
# 方式一:通过 Claude Code CLI 一键安装
claude skill install https://github.com/fritzprix/libr-agent
# 方式二:手动配置 claude_desktop_config.json
{
"mcpServers": {
"------": {
"command": "npx",
"args": ["-y", "libr-agent"]
}
}
}
# 配置文件位置
# macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
# Windows: %APPDATA%/Claude/claude_desktop_config.json
# 安装后在 Claude 对话中直接使用 # 示例: 用户: 请帮我用 智能桌面代理 执行以下任务... Claude: [自动调用 智能桌面代理 MCP 工具处理请求] # 查看可用工具列表 # 在 Claude 中输入:"列出所有可用的 MCP 工具"
// claude_desktop_config.json 配置示例
{
"mcpServers": {
"______": {
"command": "npx",
"args": ["-y", "libr-agent"],
"env": {
// "API_KEY": "your-api-key-here"
}
}
}
}
// 保存后重启 Claude Desktop 生效
A local-first desktop app for AI agents that use real tools, run in parallel, and stay under your control. Connect any LLM, add any MCP server, and let agents read files, run shells, browse the web, and automate work that actually finishes.
한국어 | 简体中文 | 日本語 | Français | Español | Deutsch | Português
LibrAgent is a local-first agent workspace built on Tauri + Rust + React. It is designed for people who want more than chat: real file access, shell execution, browser automation, MCP extensibility, and multi-agent workflows that can keep going for hours instead of falling apart after one demo loop.
Connect cloud models or local runtimes like Ollama. Import MCP servers from tools you already use. Then hand work to agents that can inspect code, edit files, run commands, browse sites, capture knowledge, and delegate subtasks without shipping your whole workflow to somebody else's VM.
Start here: Download the latest release · Jump to 5-minute onboarding · See real-world scenarios
---
ollama pull qwen3:14bLibrAgent ships with a growing library of Bundled Skills. They are not random prompts bolted on top — they are reusable operating procedures that any agent can invoke by name.
The most important day-one skills are:
| Skill | What it does |
|---|---|
system-setup | Detects and installs missing runtimes (Python, Node.js, uv) across all platforms |
mcp-installer | Registers MCP servers from npm packages, GitHub URLs, or JSON config blocks |
mcp-importer | Imports existing MCP configs from Cursor, VS Code, Windsurf, and similar setups |
specialist-creator | Designs a full agent config (system prompt, model, tools) from a role description |
crew-constructor | Scans available tools and batch-creates a matched specialist team automatically |
agent-tooling | Audits agents, detects capability mismatches, and rebalances tool assignments dynamically |
delegate | Guides parent→child session handoff with explicit context transfer and lineage tracking |
teamwork | Scaffolds the shared workspace constitution for coordinated multi-agent work |
org | Formalizes durable org identity and org-visible member hierarchy |
schedule | Creates and manages recurring scheduled task groups for unattended automation |
soul-awakening | Anchors an agent to a SOUL.md persona — tone, stance, identity |
And that's just the operator layer. LibrAgent also ships domain skills for:
deep-research-report, knowledge-distillerdocument-to-markdown, docx, pptxskill-creator, skill-deployer, playbook-creator, mcp-buildercomputer-diagnosis and other focused helpersImportant: bootstrap is a builtin capability often used alongside these skills. Bundled Skills are the reusable procedures; builtins and MCP tools are the execution substrate underneath.
---
Download the latest installer for your platform from the Releases page.
LibrAgent_0.7.32_x64-setup.exe · LibrAgent_0.7.32_x64_en-US.msiLibrAgent_0.7.32_aarch64.dmgLibrAgent_0.7.32_amd64.AppImage · LibrAgent_0.7.32_amd64.deb · LibrAgent-0.7.32-1.x86_64.rpmDeveloper Setup:
git clone https://github.com/fritzprix/libr-agent
cd libr-agent
pnpm install
pnpm tauri dev
/assistants (Role Definitions) and /playbooks (Workflow Blueprints).---
crew-constructor to auto-generate: Researcher × 3, Analyst × 1, Writer × 1delegate skillschedule---
高质量的开源MCP工具,值得关注
AI Skill Hub 为第三方内容聚合平台,本页面信息基于公开数据整理,不对工具功能和质量作任何法律背书。
建议在沙箱或测试环境中充分验证后,再部署至生产环境,并做好必要的安全评估。
✅ MIT 协议 — 最宽松的开源协议之一,可自由商用、修改、分发,仅需保留版权声明。
经综合评估,智能桌面代理 在MCP工具赛道中表现稳健,质量良好。如果你已有明确的使用需求,可以直接上手体验;如果还在评估阶段,建议对比同类工具后再做决策。
| 原始名称 | libr-agent |
| 原始描述 | 开源MCP工具:Desktop AI agent with built-in tools. No complex setup, just productivity.。⭐7 · Rust |
| Topics | ai-agentsdesktopmcpmulti-agentrust |
| GitHub | https://github.com/fritzprix/libr-agent |
| License | MIT |
| 语言 | Rust |
收录时间:2026-05-25 · 更新时间:2026-05-30 · License:MIT · AI Skill Hub 不对第三方内容的准确性作法律背书。
选择 Agent 类型,复制安装指令后粘贴到对应客户端