Shaaf's blog

A technical blog about Java, Kubernetes and things that matter

Introducing rgctl: A Code Knowledge Graph for Engineering Teams and Agents

Index once, query many times — structural facts instead of confident guesses

rgctl turns any repository into a queryable knowledge graph with pre-computed reachability, blast radius, program slicing, taint analysis, and community detection. Built in Rust for speed, designed for LLM agents and humans alike.

Posted by Shaaf's blog on Monday, September 7, 2026

I Created a Code Graph of Magento 2 for AI Agents

Tree-sitter PHP extraction, deep analysis, and agent recipes on a 25k-file monolith

How rgctl indexes Magento Open Source into a structural graph for coding agents — custom PHP Tier 1 plugin, shallow vs deep discover, honest limits on DI-heavy frameworks, and copy-paste agent workflows.

Posted by Shaaf's blog on Thursday, September 3, 2026

Blast Radius for Agent-Assisted Development on Kubernetes

Why structured impact queries matter when you pair with a coding agent

If you use Cursor or Claude Code on large repos, blast radius gives your agent bounded structural facts instead of grep guesses. A Kubernetes walkthrough for engineers: what to expect, when it helps, and when your IDE is still the better tool.

Posted by Shaaf's blog on Wednesday, September 2, 2026

Decomposing a Monolith into Microservices with Call Graph Analysis

Finding service boundaries with community detection and blast radius

A concrete decomposition of a Java EE monolith into microservices using rgctl's community detection, blast radius analysis, and CPG field mutation tracking to identify natural service boundaries and quantify coupling at each cut.

Posted by Shaaf's blog on Tuesday, September 1, 2026

Application Migration using A Graph-Driven Approach with rgctl

Using call-graph analysis to make migration decisions based on structural data

A step-by-step walkthrough of migrating a Java EE 7 CoolStore monolith from WebLogic to Quarkus, using rgctl's blast radius, community detection, and migration planning to replace guesswork with graph-driven evidence.

Posted by Shaaf's blog on Monday, August 31, 2026