Concept
Retrieval-Augmented Generation (RAG)
Don’t make the model remember — let the agent look it up.
A pattern where an agent retrieves relevant external information at run time and feeds it into the model, grounding output in real data rather than relying on what the model memorized. A core primitive of agent memory and accuracy.
Review
Not yet scored
No ClawScore review has been published yet.
Coverage
0 stories
No linked coverage yet.
Sources
0 links
Official and reference links for deeper verification.
Discovery
Noindex
Readable here, but withheld from sitemap until coverage deepens.
Retrieval-Augmented Generation (RAG) is the pattern of fetching relevant information at run time (from a document store, a database, the web) and putting it into the model’s context, so the answer is grounded in retrieved data rather than the model’s frozen training. It is how an agent stays current and cites real sources instead of confabulating.
For agents it is less a standalone technique than a building block. It underpins persistent memory (retrieving past state), tool use (a retrieval step is just another tool), and accuracy (grounding reduces hallucination). It also inherits the same trust caveat as any fetched content: retrieved text is untrusted input, so a poisoned source can steer the agent — the retrieval boundary is part of the prompt-injection surface.