Extract entity-relation triples from Obsidian notes into graph memory stores.
npx skills add https://github.com/ethan-rio/skills --skill extract-graph
Please help me install this Agent Skill. Skill: extract-graph Source: https://github.com/ethan-rio/skills/tree/main/skills/personal_agent/extract-graph Command: npx skills add https://github.com/ethan-rio/skills --skill extract-graph
This Skill solves the problem of siloed, unconnected note content that cannot be easily queried for relationships between entities, which limits the utility of personal knowledge management systems for complex research or project work.
Use the extract-graph skill to process the note 'Source -- The Bitter Lesson.md' and add its extracted entities and relationships to your graph memory store.
High-intent search queries and answers about installing and using this skill.
You can extract entity-relation triples from Obsidian notes by applying this Skill to source, wiki, and inbox items. It automatically pulls causal, temporal, and semantic relationships between entities into a standardized taxonomy using LLM-powered validation and entity deduplication.
Building a queryable knowledge graph from personal notes is best achieved by extracting structured entity-relation triples from unconnected content. This approach transforms siloed Obsidian vault data into traversable graph memory, enabling complex relationship queries for research or project work.
Yes, you can store extracted graph triples locally without an external database. The Skill integrates with Honcho graph memory stores for production use but also includes a local JSON-on-disk fallback specifically for prototyping when the schema is still being finalized.
Graph extraction works seamlessly with existing second-brain systems and PKM workflows. It processes Obsidian source notes to build traversable graph memory, allowing you to later query how different concepts, people, and events influenced each other within your personal knowledge management setup.
Idempotent triple extraction prevents duplicate entities in a knowledge graph by ensuring that processing the same Obsidian note multiple times yields consistent, non-redundant results. It combines LLM-powered validation with entity deduplication to maintain a clean and accurate graph memory store.
The limitation of using local JSON-on-disk storage for graph memory is that it is intended only for prototyping when the Honcho schema is still being finalized. For production use and complex traversable knowledge graph queries, integrating with Honcho graph memory stores is required.
cosmicdreamsIngest raw markdown notes into structured wiki pages with v2 frontmatter.
hyperb1issStore coding patterns, learnings, and tasks in a knowledge graph.
SaschaBrunnerCHQuery and visualize ArcGIS Knowledge graphs using openCypher.
chipoto69Generate a structured daily review from today's modified Obsidian vault notes.
