obsidian-ontology-sync

To extract entities and relationships from Obsidian markdown notes, configure a root SKILL.md file and run sync.py. The Skill scans your vault directories to identify entities like Person or Project and maps relationships such as works_at into an append-only ontology graph file like graph.jsonl.

Yes, you can automate Obsidian knowledge graph updates using cron scheduling. The Skill supports running extraction actions every few hours and periodically triggers analysis and feedback generation to ensure your structured ontology stays synchronized with your personal knowledge management notes.

Markdown ontology extraction detects data quality issues by computing entity and relationship statistics during the analysis phase. It identifies missing properties, such as people without email addresses, and flags orphaned entities to help maintain a clean and queryable knowledge graph.

The best way to get targeted feedback for missing information is to use the Skill's feedback generation action. It produces daily or weekly reports containing specific suggestions for missing properties, broken links, and template improvements based on the extracted ontology insights.

Yes, you need python3, jq, pyyaml, and python-dateutil installed to sync Obsidian notes into a queryable graph. These dependencies are required to execute the sync.py script for parsing markdown content and managing the append-only ontology graph output.

Yes, you can preview entity extraction before writing to the graph by requesting an extraction dry run. This allows you to see exactly which entities and relationships would be extracted from your markdown notes and written to the ontology graph file without modifying it.