I build open-source tools for people working with AI agents: better writing, reusable context, and integrations with the tools they already use.
I also host Chain of Thought, a podcast about AI infrastructure and the people building it.
Follow me here for new tools, practical examples, and what I learn maintaining them.
| Project | What you can do with it |
|---|---|
| avoid-ai-writing | Audit and rewrite common AI writing patterns with a portable agent skill and bundled detector. |
| agent-context-os | Keep project context and reviewed handoffs in Git across coding agents. See the repo for supported and experimental adapters. |
| substack-mcp | Read publication data and work on Substack drafts from an agent. Long-form posts stay drafts; publishing Notes is a separate action. |
Browse all my projects, including MCP servers, agent skills, and Mojo libraries.
Previously built developer ecosystems at Modular (acquired by Qualcomm), Galileo (acquired by Cisco), and LinearB, and worked in enterprise consulting at Microsoft. Now independent, combining open-source development, consulting, and conversations with the people shipping AI systems.
Away from that: Twilight Imperium, Age of Empires II, and a former chess nerd.
Podcast | Consulting | Writing | LinkedIn | X





