Skip to content
Back to search

haiku.rag

Agentic RAG for local and self-hosted document search: hybrid retrieval, reranking and multimodal RAG on embedded LanceDB, with Docling parsing and an MCP server

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Checked today

Check it deeper on your machine

It has changed since its scan on 11 Nov 2026.

As a Claude Code plugin, install it and then ask Claude Code:

/plugin marketplace add keithmackay/mackayi/plugin install scanskillsafety@mackayiDo a deep scan of https://github.com/ggozad/haiku.rag with scanskillsafety

Or from a terminal, clone it and run it:

git clone https://github.com/keithmackay/scanskillsafety-skill ~/.claude/skills/scanskillsafetynode ~/.claude/skills/scanskillsafety/cli.cjs --deep 'https://github.com/ggozad/haiku.rag'

It checks the exact commit: every file, its packages against OSV (api.osv.dev), how it compares with this page's scan (findsafeskills.com receives the repository's address), and, if you agree, Claude reading every file with your own Claude subscription. It never runs the code.

Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/ggozad--haiku-rag/badge.svg)](/listings/ggozad--haiku-rag)

HTML

<a href="/listings/ggozad--haiku-rag"><img src="/api/listings/ggozad--haiku-rag/badge.svg" alt="findsafeskills safety rating" /></a>

Install

i

Install instructions from the README

As written by the repo's author. findsafeskills has not verified these steps.

Python 3.12 or newer.

pip install haiku.rag        # Docling, the VoyageAI and Cohere embedders, every reranker, the TUI
pip install haiku.rag-slim   # the core, with extras chosen by you

The ingester, S3 access and model providers other than Ollama and OpenAI-compatible endpoints are extras. See [Installation](https://ggozad.github.io/haiku.rag/installation/).

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

.claude-plugin/marketplace.json · Manifest · 14 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

This file has changed since it was scanned, so this is not exactly the text that was scanned, and no line is marked. The findings are listed with the safety scan above.

  1. {
  2. "name": "haiku-rag",
  3. "description": "The haiku.rag knowledge base as Claude Code tools and a skill.",
  4. "owner": {
  5. "name": "Yiorgis Gozadinos"
  6. },
  7. "plugins": [
  8. {
  9. "name": "haiku-rag",
  10. "source": "./plugins/haiku-rag",
  11. "description": "Search, read and analyze your haiku.rag knowledge base from Claude Code."
  12. }
  13. ]
  14. }
README.md · README · 112 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

This file has changed since it was scanned, so this is not exactly the text that was scanned, and no line is marked. The findings are listed with the safety scan above.

  1. # haiku.rag
  2. [![PyPI](https://img.shields.io/pypi/v/haiku.rag)](https://pypi.org/project/haiku.rag/)
  3. [![Python](https://img.shields.io/pypi/pyversions/haiku.rag)](https://pypi.org/project/haiku.rag/)
  4. [![Downloads](https://static.pepy.tech/badge/haiku-rag-slim/month)](https://pepy.tech/projects/haiku-rag-slim)
  5. [![Docs](https://img.shields.io/badge/docs-ggozad.github.io-blue)](https://ggozad.github.io/haiku.rag/)
  6. [![Tests](https://github.com/ggozad/haiku.rag/actions/workflows/test.yml/badge.svg)](https://github.com/ggozad/haiku.rag/actions/workflows/test.yml)
  7. [![codecov](https://codecov.io/gh/ggozad/haiku.rag/graph/badge.svg)](https://codecov.io/gh/ggozad/haiku.rag)
  8. Agentic RAG that answers questions about your own documents with citations to page numbers and section headings. It runs on an embedded LanceDB database with open models through Ollama by default, so no server or API key is needed. Any provider Pydantic AI supports works in their place, and the same database can live on S3, GCS, Azure or LanceDB Cloud.
  9. Built on [LanceDB](https://lancedb.com/), [Pydantic AI](https://ai.pydantic.dev/) and [Docling](https://docling-project.github.io/docling/). Documentation: [ggozad.github.io/haiku.rag](https://ggozad.github.io/haiku.rag/).
  10. ## Features
  11. - **Ingest** PDFs, office documents, HTML, Markdown and images with Docling, in-process or on docling-serve. The stored DoclingDocument keeps headings, tables, pictures and page provenance.
  12. - **Search** with hybrid vector and full-text retrieval, optional reranking (cross-encoders, Jina, Cohere, Zero Entropy, vLLM, OpenRouter), section-aware context expansion, and image search with a multimodal embedder (vLLM, OpenRouter, VoyageAI, Cohere). Across several named databases at once.
  13. - **Answer** with the RAG capability: it searches, runs sandboxed Python over the documents for counting and aggregation, and cites page numbers and headings. Vision models receive the figures. Optional capabilities compact earlier evidence in long conversations and require every answer to declare its grounding.
  14. - **Check** a citation by drawing its chunk on the page image, from the CLI, the chat TUI or Python.
  15. - **Integrate** through the Python API, native Pydantic AI capabilities, plugin hooks that observe document writes and transform searches, an MCP server for Claude Code, Codex and Claude Desktop, and a reference web app.
  16. - **Operate** with the `haiku-ingester` service (filesystem, HTTP, S3 and WebDAV sources, a SQLite or Postgres job queue with retries, a control plane and dashboard), tags and rollback, vacuum, and `haiku-rag doctor` health checks.
  17. ## Installation
  18. Python 3.12 or newer.
  19. ```bash
  20. pip install haiku.rag # Docling, the VoyageAI and Cohere embedders, every reranker, the TUI
  21. pip install haiku.rag-slim # the core, with extras chosen by you
  22. ```
  23. The ingester, S3 access and model providers other than Ollama and OpenAI-compatible endpoints are extras. See [Installation](https://ggozad.github.io/haiku.rag/installation/).
  24. ## Quick start
  25. The default configuration uses Ollama for embeddings and answers. The [quickstart](https://ggozad.github.io/haiku.rag/tutorial/) covers the models to pull and using OpenAI instead.
  26. ```bash
  27. haiku-rag init # create the database
  28. haiku-rag add-src paper.pdf # index a file, URL or directory
  29. haiku-rag search "attention mechanism"
  30. haiku-rag ask "What datasets were used for evaluation?"
  31. haiku-rag ask "How many documents mention transformers?"
  32. haiku-rag ask "Does this figure match the spec?" --image figure.png
  33. haiku-rag chat # multi-turn chat in the terminal
  34. ```
  35. Continuous ingestion from configured sources runs as a separate service, with the ingester extra (`pip install 'haiku.rag[ingester]'`):
  36. ```bash
  37. haiku-ingester serve
  38. ```
  39. ## Python API
  40. ```python
  41. from haiku.rag.client import HaikuRAG
  42. async with HaikuRAG("knowledge.lancedb", create=True) as rag:
  43. await rag.create_document_from_source("paper.pdf")
  44. await rag.create_document_from_source("https://arxiv.org/pdf/1706.03762")
  45. results = await rag.search("self-attention")
  46. for result in results:
  47. print(f"{result.score:.2f} | p.{result.page_numbers} | {result.content[:100]}")
  48. answer, citations = await rag.ask("What is the complexity of self-attention?")
  49. print(answer)
  50. for cite in citations:
  51. print(f" [{cite.chunk_id}] p.{cite.page_numbers}: {cite.content[:80]}")
  52. ```
  53. To compose your own agent, see [Capabilities](https://ggozad.github.io/haiku.rag/capabilities/).
  54. ## MCP server
  55. ```bash
  56. haiku-rag mcp --stdio
  57. ```
  58. The server gives an assistant search, document reading and a Python sandbox over the documents. In Claude Code, the plugin registers it with a skill:
  59. ```bash
  60. claude plugin marketplace add ggozad/haiku.rag
  61. claude plugin install haiku-rag
  62. ```
  63. Codex and Claude Desktop setup is in the [MCP docs](https://ggozad.github.io/haiku.rag/mcp/).
  64. ## Examples
  65. - [Docker setup](https://github.com/ggozad/haiku.rag/tree/main/examples/docker): docling-serve, the ingester and the MCP server
  66. - [Web application](https://ggozad.github.io/haiku.rag/apps/): chat, ingestion, inspector and doctor in the browser
  67. ## Documentation
  68. - [Quickstart](https://ggozad.github.io/haiku.rag/tutorial/): install, index, chat
  69. - [Installation](https://ggozad.github.io/haiku.rag/installation/): packages and extras
  70. - [Architecture](https://ggozad.github.io/haiku.rag/overview/): how a document becomes a cited answer
  71. - [CLI](https://ggozad.github.io/haiku.rag/cli/) and [Chat and inspector](https://ggozad.github.io/haiku.rag/chat/)
  72. - [Capabilities](https://ggozad.github.io/haiku.rag/capabilities/): native Pydantic AI capabilities
  73. - [Configuration](https://ggozad.github.io/haiku.rag/configuration/): every setting, and [tuning](https://ggozad.github.io/haiku.rag/tuning/)
  74. - [Ingester](https://ggozad.github.io/haiku.rag/ingester/), [MCP](https://ggozad.github.io/haiku.rag/mcp/) and [remote processing](https://ggozad.github.io/haiku.rag/remote-processing/)
  75. - [Python API](https://ggozad.github.io/haiku.rag/python/) and [custom pipelines](https://ggozad.github.io/haiku.rag/custom-pipelines/)
  76. - [Benchmarks](https://ggozad.github.io/haiku.rag/benchmarks/) and the [changelog](https://ggozad.github.io/haiku.rag/changelog/)
  77. ## License
  78. [MIT](https://github.com/ggozad/haiku.rag/blob/main/LICENSE).
  79. <!-- mcp-name is used by the MCP registry to identify this server -->
  80. mcp-name: io.github.ggozad/haiku-rag

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.