agentsclimarketplace

R ragnar

Skill arthurgailes/r-package-skills/skills/r-ragnar

Use when code loads or uses ragnar (library(ragnar), ragnar::), implementing RAG in R, creating vector stores, embedding documents for semantic search, or building retrieval pipelinesFrom its SKILL.md

Install
npx -y skills add arthurgailes/r-package-skills --skill r-ragnar

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

  • no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
  • 23 stars23 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

2.7 KB, as published. Nobody here has run it

ragnar: Retrieval Augmented Generation in R

Overview

ragnar implements RAG in R. Create vector stores, embed documents, register retrieval tools with ellmer chat sessions. LLMs search your documents before answering.

Install: install.packages("ragnar")

References

Read references/API.md before writing code.

  • references/API.md - Complete function reference
  • references/package-docs.md - Vector store setup and usage
  • references/rag.md - RAG patterns and ellmer integration

When to Use

  • LLM needs to search your documentation
  • Implement semantic search over documents
  • Create vector database from documents
  • RAG workflows in R

When NOT to Use

  • Just need keyword search (use grep/Grep tool)
  • Documents fit in single prompt (<100k tokens)
  • Building chatbot without document search (use r-ellmer only)

Quick Reference

library(ragnar)

# Create store
docs <- read_as_markdown("docs/")
chunks <- markdown_chunk(docs)
store <- ragnar_store_create("docs.duckdb")
ragnar_store_insert(store, chunks)

# Register with ellmer
library(ellmer)
chat <- chat_openai()
ragnar_register_tool_retrieve(chat, store)

# Now chat searches docs before answering
chat$chat("What does the documentation say about...?")

# Local embeddings (Ollama)
store <- ragnar_store_create("local.duckdb",
  embed = ragnar_embed_ollama(model = "nomic-embed-text"))

Common Mistakes

IssueSolution
Embedding mismatchStore and retrieval must use same provider
Store not registeredCall ragnar_register_tool_retrieve(chat, store)
Ollama embeddings not workingStart ollama serve first
Poor retrieval qualityCheck chunking strategy, embedding model

Core Functions

Store Management:

  • ragnar_store_create(): Create vector database
  • ragnar_store_connect(): Connect to existing store
  • ragnar_store_insert(): Add documents

Document Processing:

  • read_as_markdown(): Read documents
  • markdown_chunk(): Split into chunks

Integration:

  • ragnar_register_tool_retrieve(): Register with ellmer
  • ragnar_embed_ollama(): Local embeddings

Advanced

See references/ for:

  • API.md: Complete function reference
  • rag.md: RAG workflow details
  • Package docs: Full package documentation

Integration

With ellmer: ragnar_register_tool_retrieve(chat, store) With vitals (evaluation): See r-vitals skill for RAG testing Cross-package patterns: See r-ai meta-skill

Keep looking

Skills are one crate of 613,297. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.