# Teddynote Langchain RAG > Use this skill for public-safe reasoning about RAG and LangChain pipelines, document loading, splitting, embedding, vector stores, retrievers, prompts, chains, evaluation, deployment, and troubleshooting. - Skill: `tomevault-io/teddynote-langchain-rag` (Agent Skill, multi-file: 2 files) - Install (CLI): `npx skillmds@latest add tomevault-io/teddynote-langchain-rag` - Raw SKILL.md: https://api.skillmd.com/api/skills/tomevault-io/teddynote-langchain-rag/raw - Safety review: pending (external: skill-scanner PASS, skillspector PASS) - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: AI & ML - Author: tomevault-io (https://skillmd.com/u/tomevault-io) - Updated: 2026-09-17 - Page: https://skillmd.com/skills/tomevault-io/teddynote-langchain-rag --- # Teddynote-langchain-rag Use this skill with the public-safe knowledge graph in `references/`. The private OCR source is intentionally excluded from tracked files. ## Load Order 1. Read `references/ontology.yaml`. 2. Search `references/chunks.jsonl` for topic clusters. 3. Use `references/nodes.jsonl` and `references/edges.jsonl` to connect decisions, warnings, and workflows. 4. Use `references/coverage_matrix.jsonl` and `references/query_qa.jsonl` to inspect coverage and retrievability. ## Public Boundary Do not quote or reconstruct the private OCR text. Use the graph as lossy abstraction and keep answers educational. --- > Source: [munlucky/moonshotnote-skills](https://github.com/munlucky/moonshotnote-skills) — distributed by [TomeVault](https://tomevault.io).