CaptureMind

将 Claude Desktop 对话直接整理成可视化思维导图
Turn your Claude Desktop conversations into visual mind maps — via MCP, no extra API key needed.

用 Claude 聊产品思路、创业方向、个人决策时，对话结束后思路往往还是散的。CaptureMind 让 Claude 在对话结束后直接把想法、目标、卡点、行动项结构化到一张导图里——尤其是卡点（Blocker），会被高亮显示，帮你看清自己真正卡在哪。

When you finish a deep conversation with Claude about product ideas, business strategy, or personal decisions, your thoughts are often still scattered. CaptureMind lets Claude structure them into a visual mind map — highlightingBlockersso you can finally see where you're stuck.

-

capturemindMCP 工具— Claude 自行分析对话，提取节点写入导图，无需额外 API Key
capturemindMCP tool— Claude analyzes the conversation itself and writes structured nodes to the map, no extra API key required

卡点高亮— Blocker 节点深红色高亮，一眼看出瓶颈
Blocker highlight— Blocker nodes are rendered in deep red, making bottlenecks instantly visible

自动布局— 节点按层级树状排列，不拥挤
Auto layout— Nodes are arranged in a hierarchical tree, clean and readable

双语界面— 中文 / English 一键切换，偏好自动记忆
Bilingual UI— Switch between Chinese and English with one click, preference is remembered

画布交互— 拖拽、缩放、连线、快捷键全支持
Canvas interactions— Drag, zoom, connect nodes, full keyboard shortcut support

MCP 完整控制— Claude 可直接增删节点、搜索、导出
Full MCP control— Claude can add/edit/delete nodes, search, and export without opening the browser

前置要求 / Prerequisites:Node.js 18+、Claude Desktop

# 克隆项目 / Clone git clone https://github.com/acearv09876/capturemind.git cd capturemind # 安装依赖 / Install npm install # 启动 / Start (default port 3001) npm start

浏览器打开 / Open in browser:http://localhost:3001

Claude Desktop MCP 配置 · MCP Configuration

找到 Claude Desktop 配置文件 / Locate the Claude Desktop config file:

- macOS:~/Library/Application Support/Claude/claude_desktop_config.json
- Windows:%APPDATA%\Claude\claude_desktop_config.json

{ "mcpServers": { "capturemind": { "command": "node", "args": ["/path/to/capturemind/mcp-server.js"] } } }

将/path/to/capturemind/替换为实际路径，重启 Claude Desktop。
Replace/path/to/capturemind/with the actual path, then restart Claude Desktop.

帮我用 create_topic 新建一个主题，名字叫"产品定位探讨" Create a new topic called "Product Positioning" using create_topic

② 对话结束后，让 Claude 整理 / After the conversation, ask Claude to capture it

帮我用 capturemind 把这段对话整理成思维导图，topic_id: <主题ID> Use capturemind to turn this conversation into a mind map, topic_id: <topic_id>

Claude 会自己分析对话，提取节点并写入导图。打开http://localhost:3001查看结果。
Claude will analyze the conversation, extract nodes, and write them to the map. Openhttp://localhost:3001to see the result.

帮我用 layout_topic 重新整理一下布局，topic_id: <主题ID> Use layout_topic to reorganize the layout, topic_id: <topic_id>

Manage Obsidian vaults with knowledge graph operations and AI-powered features.

Standalone Obsidian MCP server with semantic search, knowledge graph analytics (PageRank, Louvain, shortest path), and vault editing — no plugin, no REST API, works when Obsidian is closed.

AI-native knowledge vault MCP server: capture thoughts, auto-promote to structured notes, and build a knowledge graph with Louvain clustering, all from Claude Desktop.

Access and manage your Roam Research graph via its API.

Interact with TheBrain's knowledge management system using its API.

Markdown and Obsidian compatible knowledge graph.

An MCP server for interacting with the YuQue knowledge base, enabling AI assistants to perform operations on documents and information.

NVIDIA AI Developer Stack as Compressed Knowledge Graph (CKG) - 20 domains, 998 notes, agents traverse typed dependency edges instead of scanning docs.

Create notes, search, & think with your Fabric AI workspace

MCP server to interact with Routine: calendars, tasks, notes, etc.