News for hackers. Stuff that matters.

2dogsanerd

thank you...its the right lesson at the right moment ;) i learned a lot

Ask HN: Founder is claiming my 2-year RAG architecture as his AI's Featured Work

Over the last two years, I’ve built a complex Enterprise RAG system (Multi-Lane Consensus Architecture, layout-aware parsing, microservices). I open-sourced the core components (RAG_enterprise_core, smart-ingest-kit, DAUT, ingest-router even my Mail-assisten as a plugin for rag etc.) to give back to the community and build in publicYesterday, I discovered that another founder/developer and his...

( Read More... | 2 comments | 2 points | on HN )
Re:Show HN: Self-hosted RAG with MCP support for OpenClaw on Monday February 02, 2026 @10:21AM (#46857017)
Thanks! The Community Edition is intentionally a 'primitive' – single container, zero config However, I disagree on the 'maintenance tax'. ClawRAG isn't a weekend project; it's the retrieval engine extracted directly from our Enterprise RAG Core V4 system. It keeps the same connection pooling and health checks, so it is built to be 'always-on' without babying The full V4 system adds Governance (Solomon Consensus/Multi-lane validation), not basic stability. You don't need the Enterprise layer just to keep the lights on

Re: PAIO – if they implement an MCP Client, ClawRAG can serve them. But I'd argue: if you already run a host, adding a container gives you provable privacy vs. 'trust us' managed services. I prefer owning the keys AND the lock ;-)

Show HN: Self-hosted RAG with MCP support for OpenClaw

https://github.com/2dogsandanerd/ClawRag

I've been using OpenClaw to control my home server via WhatsApp, but it couldn't access my documents. Instead of uploading my private contracts to OpenAI, I built ClawRAG – a self-hosted RAG engine that connects to OpenClaw via MCP (Model Context Protocol). Now I can ask "What did the contract say about liability?" and get cited answers, not hallucinations.Most RAG systems are either too...

( Read More... | 2 comments | 2 points | on HN )
preparing to and found a pilot project with endboss pdf which im able to handle already.... but i focus on 100% quality in the db, so there is always a hitl below 100% confi...cant wait to proof im able to have a "hallucination free" rag..... main goal so far....next headache will be the update of data in the rag

Show HN: The feature gap "Chat with PDF" tuts and a regulated enterprise system

https://gist.github.com/2dogsandanerd/2a3d54085b2daaccbb1125601945ceeb

I've spent the last few months architecting a RAG system for a regulated environment. I am not a developer by trade, but I approached this with a strict "systems engineering" and audit mindset.While most tutorials stop at "LangChain + VectorDB", I found that making this legally defensible and operationally stable required about 40+ additional components.We moved from a simple ingestion script...

( Read More... | 2 comments | 2 points | on HN )

Show HN: DAUT – AI-powered documentation generator for your codebase

https://github.com/2dogsandanerd/DAUT

I built a tool that scans your code and automatically generates comprehensive documentation using local LLMs (Ollama). Features: - Scans Python/JS/TS codebases - Generates human-readable docs with AI - Resume support (skip already-done files) - Real-time progress tracking - Works 100% locally Perfect for documenting legacy code or keeping docs in sync. Built with Streamlit +...

( Read More... | no comments | on HN )

Show HN: Validated Table Extractor–Verify PDF Tables Using Docling+Vision LLMs

https://github.com/2dogsandanerd/validated-table-extractor

Hey HN,I built this because I got tired of "silent failures" in traditional PDF table extraction tools.In my day job working with financial and legal documents, tools like Camelot or Tabula often return data that looks plausible but has shifted columns or missing decimal points. In regulated environments, you can't afford to guess.I built a pipeline that treats extraction as a hypothesis to be...

( Read More... | no comments | 2 points | on HN )

Show HN: Side-by-side PDF parser comparison for RAG pipelines

https://github.com/2dogsandanerd/rag_pdf_audit

A simple tool to compare how different PDF parsers handle your documents.Shows naive parsing (pypdf) vs layout-aware parsing (Docling) side-by-side.Helps spot issues with scans, tables, and multi-column layouts before theycause problems in your RAG system.Parsers are easy to swap if you want to try alternatives.

( Read More... | 1 comment | on HN )
Hi HN, I built this "Community Edition" kit because I wanted a clean, self-hosted way to chat with my documents AND my codebase without sending data to the cloud. It's a Docker Compose setup that wires together: Docling for parsing (handles PDFs/Tables really well) ChromaDB for vector storage FastAPI backend + Simple UI Ollama for local embeddings/LLM The main feature: It separates "Code" and "Docs" into different collections with optimized chunking strategies, so you can ask "How does the auth middleware work?" and get actual code snippets back. It's open source (MIT). I stripped out the heavy async worker queues from my enterprise version to keep this kit lightweight enough for a standard VPS or local machine. Repo: https://github.com/2dogsandanerd/Knowledge-Base-Self-Hosting... I'd love your feedback on the architecture or the parsing quality!

Show HN: Smart Router Kit – Prevent "Garbage in" for RAG Using Pydantic and LLMs

https://github.com/2dogsandanerd/smart-router-kit

Yesterday I shared my Ingest Kit here. Today I'm releasing the second part: The routing logic.Problem: Blindly chunking every PDF leads to bad retrieval. Solution: A small LLM pass (Traffic Controller) that routes documents to semantic collections (Finance vs Tech) and selects chunking strategies (Table-Aware vs Standard) before ingestion.It's a simple pattern using Pydantic to force structured...

( Read More... | no comments | on HN )

Show HN: Built a tool solve the nightmare of chunking tables in PDF vs. Markdown

https://github.com/2dogsandanerd/smart-ingest-kit

Hey HN, solo dev here. After years of frustration with how LLMs handle complex documents, especially PDFs with tables, I decided to build a solution myself. My approach uses a Markdown conversion step to preserve the table structure, which seems to work surprisingly well for chunking. This little parser is the first public piece of a much larger, privacy-focused AI platform I'm building. I'm...

( Read More... | no comments | 2 points | on HN )