Break into live sessions, redact sensitive output before it reaches the model, and a brand-new documentation site. A year and a month in, with 1.4k stars and a community that keeps showing up.

Last time I had a proof of concept. This time I closed the gap: remote debugging and Windows kernel debugging, straight from VS Code, powered by the same engine as WinDbg. It's open source and on the Marketplace.

I only wanted a faster remote debugging workflow for a service running in a VM. Instead, I ended up learning DAP, wrapping WinDbg, and building a working first proof of concept debugger adapter for VS Code.

I built an autonomous AI agent that works from a PRD, verifies its own work, and ships code while I sleep. Here's Ralph.

Ollama v0.14 now speaks the Anthropic Messages API, so Claude Code can talk to local models. I tested it by building a calendar breakout game to 'reschedule' meetings. Spoiler: 'You don't need frontline models', they said...

Taking our agents beyond the built-in DevUI by exposing them as OpenAI-compatible endpoints and integrating with LibreChat - a powerful open-source chat interface.