Notes on code intelligence, developer tools, and AI-assisted software engineering.
We benchmarked AI agents on dead code detection across 60+ runs and 14 real-world repositories using Claude Opus 4.6. Graph-enhanced agents achieved 94.1% F1 with 100% precision — 156x cheaper, 11x faster, 2x better than the baseline agent alone.
Read more →
SQLite beat Kuzu and an LLM baseline in a benchmark for a predictive intent system. The article is about running the test, not the result.
A practical guide to Claude Code skills: what they are, how to write them, where they live, when not to use them, and how to package one as a plugin.
A walk through 18th-century Königsberg, an impossible puzzle, and the moment a Swiss mathematician turned a city into a graph.
Adding a verify step made my AI agent 55% dumber. Here's why asking a weaker tool to check a stronger one always backfires.
Why AI agents burn most of their tokens rediscovering a codebase, and how a small file next to each source file fixes it.
A walk through every public endpoint on the Supermodel API: what each graph represents, why we ship it as a primitive, and what you can build on top.
Why agent-written code demands a graph model, and how Supermodel builds one from your repo.