agent-bench
Compare pi, oh-my-pi and OpenCode using the same local LLM, with independent grading and inspectable reports.
Open-source tools I've built for evaluating AI coding agents and capturing thoughts, ideas and tasks with local-first software.
I build tools to explore how AI can be useful in day-to-day engineering and personal workflows. These projects focus on practical problems, local control and making the results easy to inspect. Both are open source and MIT licensed.
Explore each project to learn what it does, why I created it and where to find the source code.
Compare pi, oh-my-pi and OpenCode using the same local LLM, with independent grading and inspectable reports.
A local-first app for thoughts, ideas and tasks, with optional AI tagging, board, list and graph views, and MCP integration.
Tell me what is slowing the business down, what you have tried, and what a successful outcome looks like. I'll help you identify the most sensible next step.