Tools I build for working with agents.
Small, sharp tools I build to work better with AI agents. Desktop apps, CLIs and SDKs for code review, agent infrastructure and developer productivity. They might help you too.
Built for the era of AI agents producing 50+ file pull requests.
Each one started as something I needed for my own work. Deterministic output, no telemetry, and source you can read.
Free · 20 questions · Nothing leaves your browser
Twenty questions about what your team actually does with coding agents. Returns the autonomy level your repository can safely support, from L0 manual to L5 dark factory, and names the one area holding you there. Scored on the lowest area rather than the average, because that is what gating means.
Semantic diff layer for code review
Git gives you syntactic diffs. Diffcore adds meaning: what those changes mean, how they relate, and what order a human should read them in. Built for the era of AI agents producing 50+ file PRs.
Primitives, not frameworks. Headless infrastructure for AI agents.
A small set of reusable primitives for task state, docs-first specs, memory, coordination, and observability. Ships as a CLI, MCP server, TypeScript SDK, REST API, and monitoring dashboard. You keep the orchestration loop.