MCP工具 是 AI Skill Hub 本期精选MCP工具之一。综合评分 8.0 分,整体质量较高。我们强烈推荐将其纳入你的 AI 工具库,帮助提升工作效率。
MCP工具 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。
MCP工具 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/Mfrostbutter/ageniusdesk-ce
# 方式二:手动配置 claude_desktop_config.json
{
"mcpServers": {
"mcp--": {
"command": "npx",
"args": ["-y", "ageniusdesk-ce"]
}
}
}
# 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 MCP工具 执行以下任务... Claude: [自动调用 MCP工具 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": {
"mcp__": {
"command": "npx",
"args": ["-y", "ageniusdesk-ce"],
"env": {
// "API_KEY": "your-api-key-here"
}
}
}
}
// 保存后重启 Claude Desktop 生效
An open-source command center for n8n automation. Manage multiple instances, monitor errors, write code with AI assistance, and connect knowledge sources from a single dashboard.
Multi-Instance Management - Add any number of n8n instances by URL and API key - Switch between instances instantly - View all workflows, recent executions, and error history in one place
Error Visibility - Real-time error feed across all instances - Errors grouped by workflow, node, and error type with occurrence counts - Full node-level details and last-seen timestamps
Code Lab - Monaco-based editor for writing n8n Code-node logic - Syntax highlighting, autocomplete, and n8n node introspection - AI assistance to generate or explain code - One-click "Send to n8n" to deploy directly
AI Assistant - Chat with context from your workflows and error history - Works with OpenRouter (one key, hundreds of models), OpenAI, Anthropic, Perplexity, Groq, DeepSeek, Mistral, xAI (Grok), Together AI, local Ollama, or any OpenAI-compatible endpoint via a Custom base URL (Azure OpenAI, LiteLLM, vLLM, LocalAI, ...). Each area (Code Lab / Error Triage / Assistant) picks its own provider and model. - Function calling to query workflows, run executions, view errors - Attach MCP servers to extend the assistant with external tools - Works great with n8n-mcp by czlonkowski, an MCP server that gives the assistant deep n8n node knowledge plus workflow search, validation, and create/update tools (add it under Settings, MCP Servers) - Optional RAG over your knowledge sources via Qdrant
Knowledge Management - Register external knowledge sources (markdown files, APIs, documents) - Write and organize markdown notes with full-text search and backlinks - Folder tree structure compatible with Obsidian - Tag-based organization and navigation
Container Management - List, inspect, and manage Docker containers directly from the dashboard - Deploy new services using one-click templates - Community template library (drop a JSON file into data/templates/) - Workflow import, export, and backup
Secrets Store - Fernet-encrypted credential storage - Reference secrets as $VAR_NAME in instance API keys and MCP server configs - Resolution order: environment variable first, then encrypted store
Notifications - Inbound webhook for dashboard messages displayed as toasts - Optional Slack and Discord integrations (SLACK_WEBHOOK_URL, DISCORD_WEBHOOK_URL) - No API keys baked into the code
Insights - Execution analytics: success rates, error trends, busiest workflows - Per-instance health status
Observability (OpenTelemetry) - Embedded OTLP/HTTP receiver: point n8n's native OpenTelemetry exporter at AgeniusDesk, no external collector required - Observe view: a live trace list and parent/child execution waterfall, plus a metrics strip (executions, error rate, p50/p95, throughput) - LLM cost tracking folded into the trace layer: per-trace and per-call spend, enriched from token usage and a layered price book - Per-execution trace links from Errors and Insights
Community Modules - Install third-party modules from a GitHub repo through a two-phase inspect then install flow - A static code scan surfaces each module's declared capabilities (network hosts, filesystem write paths, subprocess) and flags risky calls before you consent - Proportional consent, a per-install audit trail, and one-click restart to activate - First module: YouTube Research (paste a link, get a structured breakdown auto-filed into your notes vault) - Heuristic review, not a sandbox; install only from sources you trust
Themes and Music - 3 built-in themes (Dark, Light, n8n) plus custom theme support - Integrated music player (Spotify, YouTube, SoundCloud, Apple Music, Tidal)
git clone https://github.com/Mfrostbutter/ageniusdesk-ce.git
cd ageniusdesk-ce
cp .env.example .env
docker compose up -d --build
Open http://localhost:3000. A setup wizard walks you through adding your first n8n instance.
Wire your n8n instance to report failures into the dashboard in real time.
One-click (recommended): once an instance is connected, open Settings > Error Handler > Install to active instance. This imports and activates the global error handler workflow (with the dashboard URL pre-filled). Then do the one step n8n requires: Settings > Workflows > Error Workflow and select the imported workflow. Repeat the install for each instance.
Manual: download the workflow JSON from the same tab (or backend/n8n_workflows/global-error-handler.json), import it via Workflows > Import from File in n8n, point the HTTP Request node at http://your-dashboard:3000/api/errors/webhook, then select it as the Error Workflow and activate it.
For production self-hosting, see docs/DEPLOY.md for: - Prerequisites and system requirements - TLS setup and reverse proxy configuration - Authentication posture (built-in or via auth proxy) - Data volume backup and recovery - Update workflow
One dashboard for every n8n instance: live stats, an execution timeline, and a real-time error feed.

Per-execution OpenTelemetry traces from n8n: a node-by-node waterfall, live metrics, and LLM cost folded into the trace layer.

| Executions & errors | Insights |
|---|---|
|  |  |
| **Code Lab** | **AI Models** |
|  |  |
| **The Harness** | **Containers** |
|  |  |
| **YouTube Research (community module)** | **The Harness, populated** |
|  |  |
| **Secrets store** | **Import / Export** |
|  |  |
Configuration is controlled via environment variables and the .env file. The full reference is at docs/CONFIG.md.
Key variables:
PORT - Dashboard port (default 3000)SECRET_KEY - Master key for secrets encryption (auto-generated if not set)ANTHROPIC_KEY, OPEN_AI_KEY, OPEN_ROUTER_KEY, OLLAMA_URL - core AI provider credentialsPERPLEXITY_KEY, GROQ_KEY, DEEPSEEK_KEY, MISTRAL_KEY, XAI_KEY, TOGETHER_KEY, CUSTOM_LLM_KEY - additional OpenAI-compatible providers (the Custom one pairs with a base URL set in Models)QDRANT_URL, QDRANT_API_KEY - Optional Qdrant RAG backendSLACK_WEBHOOK_URL, DISCORD_WEBHOOK_URL - Optional notification sinksSee docs/CONFIG.md for all options.
高质量的开源MCP工具,实现自动化控制和多实例管理
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,MCP工具 在MCP工具赛道中表现稳健,质量优秀。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 | ageniusdesk-ce |
| 原始描述 | 开源MCP工具:The command center for n8n automation operators: multi-instance management, real。⭐6 · Python |
| Topics | mcpaiautomationdashboard |
| GitHub | https://github.com/Mfrostbutter/ageniusdesk-ce |
| License | MIT |
| 语言 | Python |
收录时间:2026-06-28 · 更新时间:2026-06-28 · 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