What RAG Actually Does
RAG (Retrieval-Augmented Generation) retrieves the most relevant passages from your documents and injects them into the prompt before the model answers.
The prompt
Answer the question below using ONLY the provided context. If the context doesn't contain the answer, say "Not found in the provided documents" — do not guess. Context: [paste retrieved passages] Question: [your question]
What to replace
Swap these placeholders for your own details before running the prompt:
[paste retrieved passages]your own value[your question]your own value
Pro tip: The "answer only from context, otherwise say not found" instruction is the single most important line in any RAG prompt — it's what stops hallucinated answers.
How to use this prompt
- Copy the prompt using the button above.
- Replace [paste retrieved passages], [your question] with your own details — the more specific you are, the better the output.
- Paste it into Claude and run it.
- If the answer feels generic, add constraints: audience, length, tone, and what to avoid. That single change fixes most weak output.
Learn the technique
What RAG Actually Does
Module 1 — RAG Fundamentals · RAG & Custom AI Knowledge
Related prompts
Chunking: Why Split Size Matters
Documents are split into chunks before embedding.
ChatGPT / ClaudeCustom GPTs & Claude Projects
Both ChatGPT (Custom GPTs) and Claude (Projects) let you upload documents and set standing instructions — a no-code RAG setup.
ClaudeWriting Docs That Retrieve Well
Retrieval quality starts with document quality.
ChatGPTSQL Query Explainer
Understanding or debugging unfamiliar queries. Ideal when you inherit a legacy codebase, review a teammate's pull request, or hit a slow query you did not write. Asking for a plain-English walkthrough before the optimisation notes means you understand the intent first, so you can tell a genuine bug from a deliberate design decision.
ClaudeCode Review Pass
Pre-review pass before submitting a pull request
ClaudeTone Rewriter
Adjusting tone without rewriting from scratch
Topics
Related across the site
Reranking
Reranking takes an initial set of retrieved candidates and reorders them with a more accurate but slower model. It is one of the cheapest ways to materially improve RAG quality.
Chunking: Why Split Size Matters
Documents are split into chunks before embedding.
Chunking
Chunking splits documents into smaller passages before embedding them for retrieval. Chunk size is a key quality lever: too large dilutes relevance, too small loses the context needed to make sense.
Confidence score
A confidence score estimates how reliable an output is. Language models do not natively produce calibrated confidence, and asking a model how sure it is yields poorly calibrated self-assessment.
Debug a Real Bug With AI
Practise the debugging loop that works — evidence first, hypothesis second, fix last — on a bug you actually have. The habit this builds is refusing to accept a fix you cannot explain.
GPT vs Claude vs Gemini: An Honest Comparison
A practical comparison of the three leading AI models across writing, coding, reasoning, long documents and cost — including which one to reach for on which task, and where each genuinely falls short.