Load relevant entities from a local markdown knowledge graph into agent context.
npx skills add https://github.com/broomva/skills --skill kg
Please help me install this Agent Skill. Skill: kg Source: https://github.com/broomva/skills/tree/main/skills/knowledge/kg Command: npx skills add https://github.com/broomva/skills --skill kg
This skill solves the problem of navigating and retrieving context from a large, unstructured collection of markdown entities by using the LLM itself as a high-performance routing index.
Use the kg skill to load all entities related to the topic of agent memory into your current session.
High-intent search queries and answers about installing and using this skill.
To retrieve context from a markdown knowledge graph, the skill uses a two-tier system combining fast catalog lookups with deep body-grep fallbacks, loading relevant entities and 1-hop neighbors into your agent context for semantic inference.
The best way to navigate unstructured markdown entities for topical research is using the LLM as a high-performance routing index, automatically expanding graph connections to provide structural context beyond simple keyword matching.
Yes, you need a pre-indexed catalog file and local access to the entity directory structure to load knowledge graph entities and perform typed-edge traversal across the canonical document substrate.
Graph expansion works by automatically pulling in 1-hop neighbor entities when loading markdown documents, providing structural context that enables contradiction detection and semantic similarity analysis across your knowledge base.
Yes, pyyaml is supported as a dependency for managing a markdown-based knowledge graph, facilitating the loading and reasoning over local entities through catalog-based routing and body-grep fallbacks.
Using an LLM as a routing index for local knowledge graphs requires local access to the entity directory structure and a pre-indexed catalog file, limiting its use to pre-structured canonical document substrates rather than remote or unindexed data.
bobvarkeyStore and validate structured knowledge as a typed entity graph.
arolariuExtract atomic insights from source material into a retrievable knowledge graph.
yonatangrossOrchestrate knowledge graph memory operations with entity extraction and deduplication.
markus41Manage vector embeddings for semantic search across Pinecone, Weaviate, Chroma, and pgvector.
