---

name: retrieval-rag description: "Use R2R search, RAG, agent, completion, embedding, streaming citations, and retrieval troubleshooting workflows." disable-model-invocation: true metadata: disco-role: operating license: MIT

Retrieval RAG

Use this sub-skill when the user wants to search R2R content, run RAG or agent flows, handle streaming citations, or tune retrieval settings.

What it owns

• search modes and search settings
• RAG, agent, completion, and embedding calls
• streaming event handling and citation handling
• retrieval-focused troubleshooting

Start here

from r2r import R2RClient

client = R2RClient(base_url="http://localhost:7272")
response = client.retrieval.search(query="What is in the corpus?")
print(response.results)

Route out when the work becomes another topic

• Preparing or validating the ingested corpus: ../ingestion-documents/SKILL.md
• Building or repairing graph data: ../graph-workflows/SKILL.md
• Server/provider setup needed for RAG to work: ../server-configuration/SKILL.md
• JavaScript client usage: ../javascript-sdk/SKILL.md

Bundled assets

• references/search-rag-agent-reference.md
• references/streaming-and-citations.md
• references/troubleshooting.md
• scripts/retrieval_request_builder.py