🤖 AI Agents | 🧪 Evaluation | 🛠️ Developer Tools
I build AI agents, evaluation tools, and small tools for everyday development.
- 🧮 Modeling Workspace — Try the mathematical modeling agent.
- 🗂️ Idea Flows Site — Lay out work scattered across AI chats; @mention its plugin in ChatGPT to sync progress automatically.
- 🧭 Idea Flows Hook — Keep enrolled projects in sync from Codex, with context to return to when you switch tasks.
- 🧮 Open Math Model Agent — A local-first workspace from problem intake to code, paper, and verification.
- 🗂️ Idea Flows Site — An overview of work scattered across AI windows and chats, designed to ease the strain of switching between tasks. Status updates from your conversations, without manual maintenance.
- 🧭 Idea Flows Hook — Automatically record and sync enrolled projects from Codex, so you can find the progress and next step when you return, with less to keep in your head.
- 🔎 MathModel Verifier — Evaluate modeling reports with a fixed rubric, source quotations, and programmatic checks.
- 🧪 Verifier Benchmark — Inject known defects into reports to test whether evaluators detect them.