/ 目錄 / 演練場 / LobeHub
● 官方 lobehub ⚡ 即開即用

LobeHub

作者 lobehub · lobehub/lobehub

Design agent teams visually, run them with MCP tools, and ship via a clean chat UI — LobeHub is the team-of-agents platform.

LobeHub (formerly LobeChat) lets you compose agent personas, wire MCP tools, and run them in a polished chat surface. Strong template ecosystem; popular for personal-AI use cases.

為什麼要用

核心特性

即時演示

實際使用效果

就緒

安裝

選擇你的客戶端

~/Library/Application Support/Claude/claude_desktop_config.json  · Windows: %APPDATA%\Claude\claude_desktop_config.json
{
  "mcpServers": {
    "lobehub-mcp": {
      "command": "npx",
      "args": [
        "-y",
        "@lobehub/mcp-server"
      ]
    }
  }
}

開啟 Claude Desktop → Settings → Developer → Edit Config。儲存後重啟應用。

~/.cursor/mcp.json · .cursor/mcp.json
{
  "mcpServers": {
    "lobehub-mcp": {
      "command": "npx",
      "args": [
        "-y",
        "@lobehub/mcp-server"
      ]
    }
  }
}

Cursor 使用與 Claude Desktop 相同的 mcpServers 格式。專案級設定優先於全域。

VS Code → Cline → MCP Servers → Edit
{
  "mcpServers": {
    "lobehub-mcp": {
      "command": "npx",
      "args": [
        "-y",
        "@lobehub/mcp-server"
      ]
    }
  }
}

點擊 Cline 側欄中的 MCP Servers 圖示,然後選 "Edit Configuration"。

~/.codeium/windsurf/mcp_config.json
{
  "mcpServers": {
    "lobehub-mcp": {
      "command": "npx",
      "args": [
        "-y",
        "@lobehub/mcp-server"
      ]
    }
  }
}

格式與 Claude Desktop 相同。重啟 Windsurf 生效。

~/.continue/config.json
{
  "mcpServers": [
    {
      "name": "lobehub-mcp",
      "command": "npx",
      "args": [
        "-y",
        "@lobehub/mcp-server"
      ]
    }
  ]
}

Continue 使用伺服器物件陣列,而非映射。

~/.config/zed/settings.json
{
  "context_servers": {
    "lobehub-mcp": {
      "command": {
        "path": "npx",
        "args": [
          "-y",
          "@lobehub/mcp-server"
        ]
      }
    }
  }
}

加入 context_servers。Zed 儲存後熱重載。

claude mcp add lobehub-mcp -- npx -y @lobehub/mcp-server

一行命令搞定。用 claude mcp list 驗證,claude mcp remove 移除。

使用場景

實戰用法: LobeHub

Compose a persona agent backed by MCP tools

👤 Developers ⏱ ~15 min intermediate

何時使用: You want a domain-specific assistant (e.g. legal, finance) with the right tools attached.

步驟
  1. Compose
    Pick role + system prompt + MCP tools.✓ 已複製
    → Persona ready
  2. Test
    Chat in the LobeHub UI.✓ 已複製
    → Behaves as intended
  3. Share
    Publish to your workspace.✓ 已複製
    → Team uses it

結果: Domain-specific persona on top of MCP.

組合

與其他 MCP 搭配,撬動十倍槓桿

lobehub-mcp + dify-mcp

Use Dify for backend workflows; LobeHub for the persona UI

Combine lobehub-mcp with dify-mcp: Use Dify for backend workflows; LobeHub for the persona UI✓ 已複製

工具

此 MCP 暴露的能力

工具輸入參數何時呼叫成本
agent.invoke (see docs) Run a persona 1 call
plugin.register (see docs) Add an MCP plugin 1 call

成本與限制

運行它的成本

API 配額
Cloud plan or self-host
每次呼叫 Token 數
Per provider
費用
Free OSS; cloud plans optional
提示
Self-host for high-volume; cloud for fast start

安全

權限、密鑰、影響範圍

最小權限: lobehub:user
憑證儲存: Provider keys in account
資料出站: Configured providers
切勿授予: Share keys at workspace level if you need per-user isolation

故障排查

常見錯誤與修復

Plugin schema mismatch

Re-import the MCP plugin definition; LobeHub caches stale schemas

替代方案

LobeHub 對比其他方案

替代方案何時用它替代權衡
Open WebUIYou prefer a more chat-pure UXLess persona tooling

更多

資源

📖 閱讀 GitHub 上的官方 README

🐙 查看未解決的 issue

🔍 瀏覽全部 400+ MCP 伺服器和 Skills