RAG Implementation Guide
Guides the agent in designing retrieval-augmented generation systems that answer from your own documents
- Repository
- wshobson/agents
- Stars
- 42,095
- Licence
- MIT
- Files
- 2 files
What it does
This skill teaches the agent the design patterns for building RAG systems, where an AI model first searches a collection of documents for relevant passages and then writes an answer based on them. Once installed, the agent can recommend vector databases such as Pinecone, Chroma or pgvector, pick embedding models, choose retrieval and reranking strategies, and write example Python code using LangGraph and LangChain. It is a knowledge and code-writing guide; it does not itself index your files or connect to any database.
Great for
- Developers building a document Q&A or documentation chatbot who want a starting architecture
- Teams comparing vector databases or embedding models before committing to one
- Students learning how retrieval-augmented generation works in practice
Skip it if
- Non-developers expecting a finished search tool over their files, since the skill only provides guidance and example code
- Projects needing verified current pricing or model availability, because the model list is labelled '2026' and may go stale
- Sensitive private documents being sent to third-party embedding or vector services without a review of their data policies
Try asking
- “Design a RAG pipeline over our internal product manuals using pgvector”
- “Which embedding model should I use for multilingual support articles?”
- “Write a LangGraph example that retrieves from Pinecone and answers with Claude”
- “Explain how hybrid search and reranking would improve our document Q&A”
Install in 3 steps
- 1
Download the skill
The ZIP contains the 2 original files, the MIT licence and an attribution note.
rag-implementation.zip - 2
Add it to your agent
Claude Code (all projects)
mkdir -p ~/.claude/skills && unzip rag-implementation.zip -d ~/.claude/skills/Claude Code / Agent SDK (this project only)
mkdir -p .claude/skills && unzip rag-implementation.zip -d .claude/skills/Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.
- 3
Ask for the task
No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.
Get better results
- Ask for a specific stack (for example pgvector plus Claude) so the agent narrows the options instead of listing everything
- Check the model names and dimensions against each vendor's current docs before implementing, since the skill's list is dated
- Review the example code yourself before running it, especially the API keys and the document source you connect it to
What's inside
- references/details.md
- SKILL.md
Redistributed unmodified from wshobson/agents under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.