A desktop Markdown editor with an integrated AI assistant, built with Electron.
Write and edit Markdown files with syntax highlighting, preview, and AI-powered editing — all running locally on your machine.
Pre-built installers are available on the Releases page.
| Platform | Format |
|---|---|
| Windows | NSIS installer, portable ZIP |
- Dual-pane editor — syntax-highlighted editor with Markdown preview (`Ctrl+`` to toggle)
- File explorer — browse, create, rename, and open
.mdfiles from a folder - AI Assistant — ask the AI to edit your document, answer questions, or improve your writing
- Smart AI editing — document edits are applied silently and instantly; informational answers stream as chat
- Undo AI edits — every AI document change is reversible with the Undo button or
Ctrl+Z - Visual table builder — insert Markdown tables with a point-and-click grid editor
- File labels — colour-tag files in the explorer for quick navigation
- File backups — automatic
.markflow/backups/snapshots before every AI edit - Multi-model support — connect any OpenAI-compatible API (Groq, OpenAI, Mistral, Together AI) or run Ollama locally
- Export to PDF — one-click export via native save dialog
- Dark / light theme — persisted across sessions
- Configurable editor — font size and font family settings
git clone https://github.com/moilstack/moilstack-md.git
cd moilstack-md
npm install
npm start
npm startusesnodemon— the app auto-reloads when you change any file insrc/.
npm run packageOutput goes to the dist/ folder. Targets: NSIS/ZIP (Windows), DMG/ZIP (macOS), AppImage/DEB (Linux).
Open Settings (⚙ gear icon) → AI Models → Add Model and choose a provider type.
Any OpenAI-compatible provider works. Enter the Base URL, your API key, and a model name, then save.
| Provider | Base URL | Free? |
|---|---|---|
| Groq | https://api.groq.com/openai/v1 |
✅ No credit card |
| Google Gemini | https://generativelanguage.googleapis.com/v1beta/openai/ |
✅ Free tier |
| OpenRouter | https://openrouter.ai/api/v1 |
✅ Free models (append :free) |
| Mistral | https://api.mistral.ai/v1 |
✅ Free tier |
| Together AI | https://api.together.xyz/v1 |
✅ $1 credit on signup |
| OpenAI | https://api.openai.com/v1 |
Paid |
- Install from ollama.com/download
- Pull a model:
ollama pull qwen2.5:7b(recommended) orollama pull llama3.2 - In Settings, add a model with type Ollama and click Detect to find running models
Models below 7B parameters may not follow the document editing format reliably.
| Model | Provider | Doc Editing | Speed |
|---|---|---|---|
llama-3.3-70b-versatile |
Groq (free) | ⭐⭐⭐⭐⭐ | Fast |
gemini-2.0-flash |
Google (free) | ⭐⭐⭐⭐⭐ | Very fast |
gpt-4o-mini |
OpenAI (paid) | ⭐⭐⭐⭐⭐ | Fast |
qwen2.5:7b |
Ollama | ⭐⭐⭐⭐ | Medium |
llama3.2 |
Ollama | ⭐⭐⭐ | Fast |
- Type a prompt in the chat panel on the right and press Enter (or click the send button)
- Use Alt+Enter to insert a newline in your prompt
Ask the AI to modify your document:
Fix the grammar in this document
Add a summary section at the top
Convert this to use bullet points
Make the introduction more concise
When the AI makes a document edit:
- Changes are applied silently and instantly — no streaming wall of text
- A summary of what changed appears in the chat bubble
- Use the Undo button (↺) on the bubble to revert, or press
Ctrl+Z - Use the Re-apply button (↻) to reapply the last undone edit
- A backup is saved to
.markflow/backups/before every AI edit
Ask anything about writing, Markdown, or your document:
What's the difference between a blockquote and a code block?
How do I create a table in Markdown?
Summarise what this document is about
Answers stream directly into the chat.
Select lines in the editor before sending a prompt to scope the AI's edit to just those lines.
| Shortcut | Action |
|---|---|
Ctrl+S |
Save file |
Ctrl+Z |
Undo (AI edits first, then native undo) |
| `Ctrl+`` | Toggle Edit / Preview mode |
Ctrl+O |
Open folder picker |
Ctrl+N |
New file in current folder |
Ctrl+F |
Find & replace |
Enter |
Send chat message |
Alt+Enter |
New line in chat input |
Escape |
Close any open modal or dropdown |
Every time the AI edits your document, MoilStack .md saves a backup to:
<your-folder>/.markflow/backups/
Files are named <filename>_<timestamp>.md and the last 10 backups per file are kept automatically. Use these to recover from any unwanted AI changes.
MoilStack .md uses the OpenAI Chat Completions API format for all API-type models. Any provider that offers an OpenAI-compatible endpoint works out of the box.
| Provider | Compatible? | Notes |
|---|---|---|
| Groq | ✅ | Fully compatible, free tier |
| OpenAI | ✅ | Fully compatible |
| Together AI | ✅ | Fully compatible |
| Mistral AI | ✅ | Fully compatible |
| OpenRouter | ✅ | Fully compatible |
| Google Gemini | ✅ | Via OpenAI-compatible endpoint |
| Cerebras | ✅ | Fully compatible |
| Perplexity | ✅ | Fully compatible |
| Azure OpenAI | Different URL structure — use the deployment-specific endpoint | |
| Anthropic Claude | ❌ | Different API format — not yet supported |
| AWS Bedrock | ❌ | Requires AWS SigV4 signing — not yet supported |
| Ollama | ✅ | Via built-in Ollama type (NDJSON streaming) |
moilstack-md/
├── src/
│ ├── main/
│ │ ├── index.js # Electron main process, window management
│ │ └── ipc.js # IPC handlers (file I/O, AI requests, config)
│ ├── preload/
│ │ └── index.js # Context bridge (safe API exposure to renderer)
│ ├── assets/
│ │ ├── icon.png # App icon
│ │ └── file-icon-md.svg # .md file association icon
│ └── renderer/
│ ├── index.html # App shell
│ ├── index.js # Main renderer entry point
│ ├── styles.css # All styles
│ ├── aiConfig.js # AI model settings UI
│ ├── aiService.js # AI request dispatcher
│ ├── chatPanel.js # AI chat panel
│ ├── editorCore.js # Editor logic, undo, ruler
│ ├── fileTreeManager.js # File explorer tree rendering
│ ├── fileOperations.js # Create, rename, delete files/folders
│ ├── markdownRenderer.js # Markdown → HTML preview
│ ├── saveManager.js # Save / dirty-state tracking
│ ├── storageManager.js # localStorage / IPC persistence
│ ├── tableBuilder.js # Visual table builder modal
│ ├── exportService.js # PDF export
│ └── themeManager.js # Dark / light theme
├── build/
│ ├── installer.nsh # NSIS installer customisation
│ └── make-file-icon.js # Icon generation script
├── assets/ # Screenshots and marketing assets (add here)
├── CHANGELOG.md
├── package.json
└── README.md
Issues and pull requests are welcome. For significant changes, please open an issue first to discuss what you'd like to change.
See CHANGELOG.md for version history.
MIT — see LICENSE for details.
The project name, logo, and visual branding assets are not open source. See BRANDING.md for the Trademark & Branding Policy.