16:[[44814082,{"by":"srijanshukla18","descendants":12,"id":44814082,"kids":[44815921,44824482,44816133,44815916,44820245,44816158,44816124,44822490],"score":5,"text":"Hi HN,

I built XRAY MCP after discovering that AI assistants were scanning my projects with plain grep and guessing. I tried direct tree-sitter integration and language servers; both felt heavy for a lightweight tool.

ast-grep hit a sweet spot: syntax-aware search in a single binary. XRAY MCP wraps it behind three endpoints—map, find, impact—so a model (or human) can answer questions like “what breaks if I change this function?” on demand.

It’s stateless, supports Python/JS/TS/Go, and installs quickly.

Repo: https://github.com/srijanshukla18/xray

Would love to know your thoughts!","time":1754497189,"title":"Show HN: XRAY MCP – AST-grep wrapped in a tiny server for code-aware AI","type":"story","url":"https://github.com/srijanshukla18/xray"}],[44624003,{"by":"srijanshukla18","descendants":3,"id":44624003,"kids":[44624259,44626254],"score":8,"text":"Built this as a toy project to understand knowledge graphs by tackling a real problem: traditional RAG fails badly on legal documents because it misses interconnections between sections.

The system actually combines both approaches on every query - gets semantic matches via TF-IDF, retrieves structural relationships from Neo4j, then feeds both contexts to OpenAI for comprehensive answers.

Used the Indian Income Tax Act as test data since legal documents have natural graph structures. Queries like "What sections reference Section 80C?" get both the reference network AND content explanations.

Full transparency: includes some AI-assisted code as I was learning Neo4j/graph concepts, but the hybrid architecture and problem framing are mine.

Tech stack: Python, Neo4j, OpenAI API, scikit-learn (TF-IDF), numpy. Docker + Makefile for easy setup.

Would love feedback on this pattern for other structured documents.","time":1753009778,"title":"Show HN: Hybrid Knowledge Graph and RAG for Legal Documents (Learning Project)","type":"story","url":"https://github.com/srijanshukla18/ita-kg"}]]

1.	Show HN: FFlags – Feature flags as code, served from the edge(fflags.com) 42 points by tushr 332 days ago \| 36 comments
2.	Replit replaced thanos with this(last9.io) 6 points by tushr 2 years ago \| 0 comments
3.	Show HN: A simple understanding of High Cardinality(highcardinality.com) 2 points by tushr 2 years ago \| 0 comments
4.	2 years ago \| discuss

user:	tushr
created:	December 11, 2021
karma:	9