Can Claude Code migrate Notion to Obsidian?
- 07/09/2025
Description
This video demonstrates how to efficiently migrate a large Notion workspace to Obsidian using Claude Code, an AI automation tool. It outlines a five-step process, emphasizing a structured “training system” (Claude.md file) to teach Claude how to auto-organize thousands of files, significantly reducing the manual effort typically associated with such migrations. The creator showcases how Claude Code handles complex file sorting, tagging, and organization, transforming chaotic imported files into a perfectly structured Obsidian vault.
Best Ideas
1. AI-Powered Bulk Migration: Claude Code can automate the migration and organization of thousands of Notion files into Obsidian, saving days or weeks of manual work.
2. AI “Brain” (`Claude.md`): Creating a `Claude.md` file to define specific vault structure, organizational rules, tag taxonomy, and critical plugins acts as an essential training manual for Claude Code, allowing it to understand and execute tasks according to personal preferences.
3. Milestone-Based Migration Tasks (`migration_task.md`): Breaking down the migration into manageable, bullet-point milestones within a `migration_task.md` file enables a phased approach, making the process less daunting, easier to monitor, and allowing for pauses and resumptions.
4. Progress Tracking (`migration_review.md`): Utilizing a `migration_review.md` file to document progress, changes, achievements, and discrepancies at each milestone provides a clear audit trail and aids in quality assurance.
5. Tracking Tags: Applying a specific tag (e.g., `my_notion_note`) to all migrated Notion files allows for easy identification, filtering, and review of all AI-processed content throughout and after the migration.
6. Hybrid Approach: While AI automates the bulk of the work, the video highlights the importance of manual intervention for specific cases (e.g., very long file paths, final review) and providing clear instructions to the AI.
7. Pre-Migration Planning: Prioritizing and categorizing Notion content (e.g., “don’t need,” “maybe need,” “definitely need”) *before* starting the migration helps guide the AI and streamline the organization process.
Tools and Resources
- AI Automation: Claude Code (requires Claude Pro membership @ $20/month)
- Knowledge Management Systems: Notion (source), Obsidian (destination)
- Core Development Tools:
- Node.js (foundation)
- Git (version control)
- VS Code (interface for Claude Code)
- Obsidian Plugin: Importer (community plugin for importing Notion .zip files)
- Files/Techniques:
- `Claude.md`: AI training manual, context for Claude Code.
- `migration_task.md`: Blueprint for migration tasks, broken into milestones.
- `migration_review.md`: Progress tracking hub, summary of changes.
- Specific tagging strategy (e.g., `my_notion_note` for tracking migrated files).
- NPM commands (for installing Claude Code).
- Bash commands & Python scripts (executed by Claude Code for file operations).
- Obsidian Dataview (implied for filtering and showing folder properties via queries).
- Manual file moving/renaming for edge cases (e.g., long file paths).
- Backup strategy (make a copy of your vault before starting).
Key Learning Points
- AI for KM is Powerful but Requires Guidance: AI tools like Claude Code can drastically automate complex knowledge management tasks, but they require clear, structured instructions and contextual information (e.g., via `Claude.md`) to operate effectively according to your specific needs.
- Structured Processes are Key to Complex Migrations: Even with advanced AI, breaking down a large migration into logical milestones and having a dedicated plan (`migration_task.md`) is crucial for managing complexity, tracking progress, and ensuring a successful outcome.
- Continuous Review and Quality Assurance: Don’t set and forget. Regularly reviewing the AI’s output at each milestone and documenting progress (`migration_review.md`) is essential for catching errors, making adjustments, and ensuring the final result meets your standards.
- Privacy and Backups are Paramount: Always be mindful of privacy when feeding data to AI and, more importantly, always create a full backup of your vault before initiating any large-scale, automated changes.
- Time vs. Perfection Trade-off: AI significantly reduces manual labor time but might not achieve 100% perfection. Be prepared for some manual follow-up work (e.g., refining YAML frontmatter, deleting leftover files, minor categorization tweaks).