This document outlines potential future features for remarkable-mcp. These are ideas under consideration, not commitments.
Track progress: See open enhancement issues on GitHub.
Write Support (#24)
Write tools (upload, create folders, move, rename, delete) are now enabled by default — pass --read-only to disable them. Further ideas under consideration:
- Create documents — Create new notebooks or upload PDFs
- Sync from Obsidian — Push markdown notes to reMarkable as PDFs
- Template support — Apply templates when creating notebooks
- Folder management — Create, rename, move folders
Write support requires careful consideration of:
- Sync conflicts with reMarkable's own sync
- Data safety and backup
- API stability
Additional OCR Providers (#25)
Google Vision works well, but more options would be valuable:
| Provider | Status | Notes |
|---|---|---|
| Google Vision | ✅ Implemented | Excellent handwriting recognition |
| Tesseract | ✅ Implemented | Offline fallback, poor for handwriting |
| Microsoft Azure | 🔮 Planned | Competitive handwriting OCR |
| Mistral | 🔮 Planned | Open-weight models with vision |
| Claude Vision | 🔮 Possible | Direct integration with Claude |
| Local LLaVA | 🔮 Possible | Fully offline, privacy-focused |
The goal is BYOK (Bring Your Own Key) — let users choose their preferred provider.
Enhanced Search (#26)
- Full-text indexing — Index all documents for instant search
- Semantic search — Find documents by meaning, not just keywords
- Cross-document search — Search annotations across your entire library
Deep integration with Obsidian vaults:
- Bi-directional sync — Notes flow between reMarkable and Obsidian
- Link resolution — reMarkable documents as Obsidian attachments
- Daily notes — Sync reMarkable journals to Obsidian daily notes
Export Features (#27)
- PDF export — Export notebooks as PDFs
- Markdown export — Convert notebooks to markdown
- Batch export — Export entire folders
Have an idea? Open an issue on GitHub with the enhancement label.
Popular requests we're tracking:
- Handwriting-to-text conversion — Beyond OCR, actual handwriting recognition
- Tag support — Organize documents with tags
- Favorites — Quick access to frequently-used documents
- Version history — Access previous versions of documents
Performance (#28)
- Parallel resource registration — Faster startup for large libraries
- Incremental sync — Only fetch changed documents
- Persistent cache — Cache OCR results across sessions
Reliability (#29)
- Automatic reconnection — Recover from dropped SSH connections
- Retry logic — Handle transient API failures
- Health checks — Proactive connection monitoring
- TypeScript types — Full type definitions for MCP clients
- Example integrations — Sample code for common use cases
- Plugin system — Extensible architecture for custom features
Interested in implementing any of these features? We welcome contributions!
- Check existing issues for the feature
- Open a discussion if it's a major change
- Fork, implement, and submit a PR
See Development Guide for setup instructions.
Some things we're explicitly not planning:
- reMarkable firmware modifications — We work with the official software
- Bypassing DRM — We respect content protection
- Subscription circumvention — Cloud API requires Connect subscription
- Real-time sync — We're a query tool, not a sync service