Compile sources into a personal knowledge base with LLMs.
npx skills add https://github.com/sbsbhong/diving-app --skill karpathy-llm-wiki-sbsbhong
Please help me install this Agent Skill. Skill: karpathy-llm-wiki Source: https://github.com/sbsbhong/diving-app/tree/main/.agents/skills/karpathy-llm-wiki Command: npx skills add https://github.com/sbsbhong/diving-app --skill karpathy-llm-wiki-sbsbhong
This Skill addresses the challenge of building and maintaining a comprehensive, personal knowledge base using Large Language Models (LLMs).
Run the ingest command to add new sources to your knowledge base.
High-intent search queries and answers about installing and using this skill.
To build a personal LLM knowledge base, you need to ingest raw sources, compile them into articles, and define a schema file. This Skill automates fetching, organizing, and archiving synthesized answers for efficient querying.
Organizing research material with LLMs involves compiling sources into structured articles and querying the synthesized knowledge base. This Skill provides directories for raw sources, compiled articles, and schema definitions to maintain information organization.
You can ingest and compile sources into a knowledge base by running the ingest command to fetch and structure new information. The system requires dedicated directories for raw sources and compiled articles to manage the workflow.
Yes, a schema definition file is required to maintain an LLM knowledge base. It ensures quality checking and consistency when compiling raw sources into structured articles and archiving synthesized answers.
You can check the quality and consistency of your LLM knowledge base by linting the compiled articles. This Skill supports quality checking to ensure the ingested sources meet the standards defined in your schema file.
Querying a personal knowledge base with LLMs allows you to search and retrieve information from compiled articles. The Skill supports archiving the synthesized answers generated from your queries for future reference.
ccomkhjIngest raw sources into immutable raw/ files and compile structured wiki articles.
zilliztechDecompose questions, retrieve across documents, and synthesize cited answers.
paxlabs-incIngest raw sources into an interlinked markdown knowledge base with provenance tracking.
zzafergokDesign and build RAG systems grounding LLM outputs in external knowledge sources.
