README file from
GithubSuperMind - Enhanced Mind Map Plugin for Obsidian
A powerful, stable mind mapping plugin for Obsidian
Interactive Editing • AI Smart Expansion • Multiple Export Formats • Enterprise-Grade Stability
✨ Key Highlights
🎯 Complete Mind Mapping Experience
- Native Obsidian Integration: Seamlessly integrates into your Obsidian workflow
- Bidirectional Markdown Sync: Real-time conversion between mind maps and Markdown
- 13 Professional Themes: Professional themes + dopamine color schemes for various scenarios
- Rich Text Support: Tables, Callouts, quotes, code blocks, embedded content
🤖 AI-Powered Enhancement
- Unified AI Engine: Powered by PiAI, supporting Google Antigravity, OpenAI, Anthropic, Gemini, Deepseek, CodeBuddy, OpenAI-compatible endpoints, and more
- OAuth / API Key Sign-in: Auth modal with two authorization modes, credentials stored securely; official Antigravity Loopback + PKCE authorization flow
- Native Web Access Subsystem: Zero-dependency
web_searchandweb_fetchwith SSRF protection and Auto routing (DuckDuckGo, Bocha, Tavily, Jina) - Multi-Provider AI Image Generation: Independent
ImagesModelsruntime supporting Google Antigravity, OpenRouter (52+ image models), and OpenAI-compatible / SiliconFlow endpoints - WeChat Publishing Skills Suite: Built-in tech writing, KM articles, typography formatting, and one-click draft box publishing
- Universal AI Copilot: Direct assistant invocation without pre-selecting skills, dynamically adapting as tools load skills
- Skill Task Drawer: Non-blocking docked drawer with collapsible floating bubble, multi-source context injection (selection feeding, 0-token
@file mention), and multi-turn refinement - Karpathy Wiki Agent (
wiki): Built-in full-vault knowledge compilation, note ingestion, and incremental absorption - Multimodal Vision: Dynamic vision model resolution with automatic image attachment extraction from notes and mind maps
- Pi Agent Tool Calling: read / edit / write / find / grep / ls / bash / web_search / web_fetch / generate_image — AI directly reads/writes notes, searches the web, and generates illustrations
- Native XMind Creator Skill: Built-in 15 core skills, allowing AI to directly write
.xmindfiles with boundaries, summaries, callouts, and relationships - Comprehensive Internationalization: Task drawer, auth modal, settings, and all 15 built-in skills fully localized in English, Simplified Chinese, and Traditional Chinese
- Smart Expansion: One-click child node generation, deep analysis, content optimization
- Insight System: Flomo-like AI insights with multi-dimensional analysis, directly analyzing
.xmindfiles alongside notes - Custom Prompts: Fully customizable AI interactions
🏗️ Enterprise-Grade Stability
- Unified Resource Management: Centralized resource lifecycle control
- Zero Memory Leaks: Complete lifecycle management and automatic cleanup
- Async Safety: Prevents race conditions accessing destroyed objects
- Optimized Performance: Efficient rendering and memory usage
🎨 Rich Export Options
- Image Export: PNG, JPEG, SVG with watermarks and author info
- Document Export: Selected content export, 12 professional templates
- XMind Integration: Bidirectional conversion with XMind files (Desktop)
🚀 Core Features
1. Mind Map Editing
Interactive Editing
- ✅ Create and edit mind maps directly in Obsidian
- ✅ Support for drag-and-drop, hotkeys, context menu
- ✅ Real-time preview, WYSIWYG
- ✅ Complete undo/redo history
Layouts and Styles
- 4 Layout Directions: Center, Right, Left, Clockwise
- 13 Themes: 7 professional + 6 dopamine color schemes
- Custom Color Groups: Smart node connection color system
- Responsive Canvas: Auto-adapts to content size
Rich Text Rendering
| Content Type | Supported Features |
|---|---|
| Markdown Formatting | Bold, italic, highlight, strikethrough, code |
| Tables | Full table rendering and editing |
| Callouts | Multiple Callout types, theme-adaptive |
| Block Quotes | Block quotes, text block display |
| Code Blocks | Syntax highlighting, rendered as child nodes |
| Embedded Content | Canvas, PDF++, Eagle, Excalidraw |
| Internal Links | Support for internal link navigation |
| Footnotes | Superscript rendering, content protection |
| Task Lists | Clickable checkboxes |
2. View Enhancements
Outline View
- Tree structure display
- Real-time search and quick navigation
- Expand/collapse support
- Synchronized scrolling with main view
Map Overview
- Thumbnail minimap
- Visual navigation
- Quick jump
- Real-time updates
Quick Settings
- One-click theme switching
- Adjust layout direction
- Modify color groups
- Background color picker
3. AI Features
Node AI Expansion (Plus Feature)
- Generate Ideas: AI brainstorming
- Generate Analysis: Deep content analysis
- Detailed Expansion: Refine node content
- Practical Applications: Practical scenario suggestions
- Custom Prompts: Fully customizable
Context Menu AI
- Popup with streaming output
- Support for editing and insertion
- Smart Callout type selection
- Regenerate and replace functionality
File AI Analysis
- Document analysis and summary
- Keyword extraction
- Streaming output to new file
- Custom save path
AI Insight System
- Multi-dimensional perspective analysis
- Customizable built-in perspectives
- Backlink expansion depth configuration
- Persistent data storage
AI Skills System (Plus Feature)
- Extensible AI workflows defined via SKILL.md, following the Agent Skills open standard
- Multiple output formats: markdown, mermaid, excalidraw, canvas, base, and native
.xmindmind maps - Built-in 7 core skills, including the new
xmind-creatorskill - Auto-scans vault
skills/folder to discover skills, loading them on demand via progressive disclosure - ReAct agent tool loop for complex multi-step tasks
- Streaming output, real-time file writing; automatically embeds artifact wikilinks (
![[artifact]]) into source notes - Smart content extraction, strips AI explanatory text
Skill Task Drawer & Multi-Turn Refinement
- Non-Blocking Drawer: Operates smoothly without a blocking modal backdrop, allowing uninterrupted reading and editing
- Multi-Turn Session Refinement:
SkillAgentSessionretains context across follow-up prompts for iterative polishing - Collapsible Floating Bubble: Minimize drawer into a bottom-right pill badge with a breathing pulse dot indicator
- Real-Time Activity Feed: Live visualization of tool executions (read/edit/write/bash), execution turns, and token usage
- Artifact Card: Detects the final deliverable and opens it in an Obsidian leaf with one click
Multimodal Vision & Link Resolution
- Dynamic Vision Resolution: Automatically verifies model vision capabilities across providers and custom endpoints
- Rich Content Extraction:
LinkResolverparses wikilinks, block refs, and raster images (![[image.png]],[alt](url)) - Multimodal AI Integration: Selected text, mind map nodes, and multi-document insights automatically deliver images to vision models
- Privacy Control: Enable
blockImagesin settings to disable image uploading at any time
Pi Agent Tool Calling
- Built-in
read/edit/write/find/grep/ls/bashtools editperforms batched targeted replacements instead of rewriting whole fileswritecompiles Markdown directly into native binary.xmindmind mapsgrep/findregex search across the vault; multi-document insight retrieves on demand- Physical truncation protection (
truncateOutput) protects the context window - Optional "Require Tool Confirmation" (
requireToolConfirmation) setting before modifying files or executing bash
File & Folder Context Menu AI
- Multi-select files or right-click folders in the file list to launch AI insights directly
- Folder right-click batches both Markdown documents and
.xmindfiles - Direct in-memory extraction of
.xmindoutlines and images for AI analysis
Unified Save Path Management
- Three modes: custom path / vault root / source file directory
- Shared across all AI features
- Auto-creates directories when they don't exist
AI Stop Button
- Both AI Insight and Skills execution support interruption
- Mobile progress indicator integrates stop button
- Mind map node AI expansion also supports stopping
Plus Feature Licensing
- Trial Period: 7 days free trial for all Plus features
- Licensing: Obtain registration code through sponsorship
- Scope: Node AI expansion, context menu AI, file AI analysis
- Free Features: All other features remain unrestricted after trial
4. Export and Publishing
Image Export
Supported Formats: PNG, JPEG, SVG
Features:
- Adjustable image width
- Custom author info (avatar, name, extra text)
- Watermark settings (text/image, transparency, rotation)
- 12 professional templates
- Mobile share menu integration
Document Export
Features:
- Selected content export
- Full document export
- Card summary display
- Filename and date control
- Export preview modal
XMind Integration (Desktop)
- Markdown ↔ XMind conversion
- XMind file preview (thumbnails)
- Folder auto-sync
- Open XMind in tabs
5. Collapse State Persistence
- Uses
<!--c-->comment markers - Compatible with obsidian-workflowy-plugin
- Markdown view synchronization
- Auto-save state
📱 Platform Support
Desktop (Full Support)
- ✅ Windows, macOS, Linux
- ✅ All features available
- ✅ XMind integration
- ✅ File system operations
Mobile (Basic Support)
- ✅ iOS, Android
- ✅ Mind map viewing and editing
- ✅ Touch operations
- ✅ Image export (share menu)
- ✅ Tab/Enter virtual buttons
- ⚠️ Not Supported: XMind conversion, file sync
📖 User Guide
Quick Start
1. Create Mind Map
Method 1: Ctrl+P → Search "Create new mind map"
Method 2: Right-click folder → "New mind map"
2. Open Existing File
Method 1: Right-click Markdown file → "Open as mind map"
Method 2: Command palette → "Toggle markdown or mindmap mode"
Keyboard Shortcuts
| Shortcut | Function |
|---|---|
Shift+F2 |
Edit node |
Shift+Insert |
Insert child node |
Alt+Shift+Enter |
Add sibling node / End editing |
Shift+Delete |
Delete node and children |
Escape |
Cancel editing |
Alt+Shift+S |
Select node text |
Alt+Shift+D |
Move next siblings as children |
Alt+Ctrl+Shift+D |
Move all siblings as children |
Alt+Shift+J |
Join with node below |
Alt+Ctrl+Shift+J |
Join as citation with node below |
| Shortcut | Function |
|---|---|
Alt+Shift+↑ |
Move node up |
Alt+Shift+↓ |
Move node down |
Alt+Shift+← |
Move node left |
Alt+Shift+→ |
Move node right |
| Shortcut | Function |
|---|---|
Alt+↓ |
Expand one level |
Alt+↑ |
Collapse one level |
Alt+PageDown |
Expand one level from max displayed |
Alt+PageUp |
Collapse one level from max displayed |
Ctrl+Shift+Space |
Toggle expand/collapse |
| Shortcut | Function |
|---|---|
Alt+Shift+B |
Bold |
Alt+Shift+I |
Italic |
Alt+Shift+H |
Highlight |
Alt+Shift+2 |
Strikethrough |
Alt+Shift+L |
Remove line breaks |
| Shortcut | Function |
|---|---|
Alt+Shift+C |
Copy node |
Alt+Shift+V |
Paste node |
Alt+Shift+Z |
Undo |
Alt+Shift+Y |
Redo |
Alt+Ctrl+Shift+Z |
Replace with previous text |
Alt+E |
Center current node |
Alt+Shift+E |
Center entire map |
Ctrl++ / Ctrl+Scroll↑ |
Zoom in |
Ctrl+- / Ctrl+Scroll↓ |
Zoom out |
Ctrl+0 |
Reset zoom |
AI Configuration
-
Set Up AI Service
- Open plugin settings → AI Service Configuration
- Select a provider (OpenAI, Anthropic, Gemini, Deepseek, CodeBuddy, etc.)
- Authenticate via API Key or OAuth, then pick a model
-
Use AI Expansion
- Click the 🧠 button on a node
- Or
Ctrl+Double-clickthe node - Select expansion method
-
Custom Prompts
- Plugin settings → AI Custom Prompts
- Use
{{nodeContent}}placeholder - Takes effect immediately after saving
🛠️ Installation
Using BRAT (Recommended)
BRAT (Beta Reviewers Auto-update Tester) allows you to install and automatically update plugins directly from GitHub.
- Install the BRAT plugin from Obsidian Community Plugins
- Enable BRAT in Settings → Community plugins
- Open BRAT settings and click "Add Beta plugin"
- Enter the repository URL:
https://github.com/springrain1/obsidian-xmind-plugin - Click "Add Plugin" and BRAT will install SuperMind automatically
- Enable SuperMind in Settings → Community plugins
Tip: BRAT will automatically check for updates and notify you when a new version is available.
Manual Installation
- Download the latest release from Releases
- Extract to:
<vault>/.obsidian/plugins/obsidian-xmind-plugin/ - Restart Obsidian
- Enable plugin in Settings → Community plugins
🔄 Latest Version
v3.3 - Google Antigravity, Native Web Access, Unified Image Generation & Skill Pipelines
- 🌐 Google Antigravity Provider & OAuth: Desktop local loopback (port 51121) + PKCE authorization code grant with support for
gemini-3-pro,gemini-3-flash,claude-4-6-sonnet, and more - 🔍 Zero-Dependency Native Web Access: Built-in
web_searchandweb_fetchwith enterprise SSRF defense and Auto routing (DuckDuckGo, Bocha, Tavily, Jina) - 🎨 Unified Multi-Provider AI Image Generation: Independent
ImagesModelsruntime supporting Google Antigravity, OpenRouter (52+ image models), and OpenAI-compatible / SiliconFlow endpoints - ⚡ 0-Token Drawer Slash (
/) Skill Autocompletion: Type/in the drawer input to instantly search and autocomplete all 15 enabled skills with keyboard navigation - 🔗 Multi-Skill Sequential Pipelines: Multi-turn context handover protocol seamlessly carrying assets across stages (e.g. WeChat "Draft ➔ Format ➔ Publish") in a single session
- 🧠 CodeBuddy Deep Reasoning & Multimodal Vision Hardening: Dynamic
reasoning_effortbinding, contract-compliant thinking level mapping, unrestricted multimodal vision, and settings persistence fixes - 📱 Modernized WeChat Publishing Skills Suite: 4 core WeChat publishing skills aligned with
$VAULT_PATHconventions and secure credential isolation - 🌍 Comprehensive Internationalization: Full elimination of hardcoded strings across drawers, modals, and settings; native localization for all 15 skills across EN, ZH-CN, and ZH-TW
v3.2 - Universal Copilot, Context Injection & Wiki Agent
- ⚡ Universal AI Copilot Drawer: Launch task drawer without a designated skill; dynamically auto-detects and binds skill titles/icons when tools read skill definitions
- 📝 Selection Snippet Injection: Highlight text across notes and right-click "Send selection to AI Drawer" with precise 1-based line bounds and >6000-char read pointer fallback
- 🔍 0-Token Local
@File Mentions: Type@in the input bar to search and attach Vault files (.md, .json, .csv, .xmind) via native fuzzy match with zero Token overhead - 📚 Knowledge Wiki Agent (
wiki): Built-in 8th skill based on Karpathy LLM-Wiki specifications with automated asset syncing and quick action suggestions - 🛡️ Unified Attachment Row: Unified 16×16px remove buttons and layout overflow protection across image thumbnails and context capsules
- ⚙️ Hardened Parser Engine: Fully supports YAML block scalars (
|,>,>-), preserving internal line breaks and fixing delimiter edge cases
v3.1 - Skill Task Drawer, Multimodal Vision & XMind Interoperability
- 🎯 New Skill Task Drawer: Non-blocking drawer UI, multi-turn follow-up refinement, live activity feed, and deliverable artifact card
- 🫧 Collapsible Floating Bubble: Dock drawer into a sleek bottom-right pill badge with a running pulse indicator
- 👁️ Global Multimodal Vision: Dynamic model vision resolution and intelligent image extraction via
LinkResolver - 🧠 New native
xmind-creatorskill (expanding default skills to 7) generating full.xmindfiles from Markdown with advanced components - 📁 Seamless
.xminddocument support in right-click menus and multi-document AI insights - 🛡️ New tool confirmation safety setting, standalone
truncateOutputprotection, and automatic artifact wikilink embedding - 🧹 Removed deprecated XMind previewer and legacy diff code (-4300+ lines), full localization across EN, ZH-CN, and ZH-TW
v3.0 - PiAI Engine & Pi Agent Toolchain
- 🤖 AI core fully migrated to the PiAI engine; unified providers with OAuth / API Key auth and secure credential storage
- 🔑 New AI auth modal and CodeBuddy provider (device-code OAuth, JWT, streaming responses)
- 🛠️ New Pi Agent toolset: read / edit / write / find / grep / ls / bash
- 🔁 Skill executor refactored into a ReAct agent loop; XML skill list injected into the system prompt
- 📂 Multi-select files and right-click folders launch AI insights directly
- 🐛 Lifecycle management prevents memory leaks, legacy settings auto-migrate, build compatibility improved
v2.8 - Mind Map View Enhancements & Format Improvements
- ✨ Mind map view now renders node tables with support for child nodes under table nodes
- ✨ Fixed rendering of Callout, table, code block, and quote block components; block-level components preserved as renderable child nodes
- 🐛 Fixed blank line removal between components when writing back to Markdown, fixed note blank line loss
- 🐛 Fixed content hierarchy misalignment after adding new nodes and converting to Markdown
- 🐛 Fixed blank line issues between same-level unordered lists after list conversion
- 🤖 New skill: xmind-mdoutline-generator - Extract mind map outline from original text, automatically add XMind-specific components
v2.7 - XMind Conversion Settings & Bug Fixes
- ⚙️ Three save path modes for XMind → Markdown attachments: Obsidian default/original file location/custom path
- 🐛 Fixed mouse drag inertia issue in canvas overview
- 🐛 Added context window token count configuration option in AI service settings
- 🐛 Fixed notes and
---separator loss issues - 🐛 XMind→Markdown note formatting optimization using more stable
<strong>/<em>syntax
v2.6 - Batch Conversion & Comprehensive Format Support
- 🔄 File list supports folder right-click or Alt multi-select for batch XMind ↔ Markdown conversion
- ✨ Comprehensive format support: colors, notes, tags, summaries, boundaries, collapse states, formulas, images, floating topics, multiple sheets
- ✨ New command to convert between "Level 1-3 headings + unordered lists" and "All unordered lists"
v2.5 - Settings Page Redesign
- ✨ Two-level tab navigation layout for settings (General/AI/License), no more endless scrolling
- ✨ Dopamine purple gradient navigation bar, light/dark theme compatible
- ✨ Lazy-loaded tab content panels with tab state memory
- 🐛 Fixed BASE view AI Insight unable to retrieve filtered documents (adapted to Obsidian's updated DOM structure)
📖 See Full Changelog
💬 Feedback & Support
If you encounter any issues or have suggestions:
- Submit an Issue on GitHub
- Describe the problem in detail with steps to reproduce
🙏 Acknowledgments
- Obsidian Community
- XMind Team
- All users and supporters
Made with ❤️ for Obsidian