I build AI developer tools, AI quality engineering tools, and local-first apps that solve real problems.
| Direction | Focus |
|---|---|
| AI Developer Tools | AI coding workflow, agent tooling, usage & observability |
| AI Quality Engineering | RAG / Agent / MCP / prompt / OCR quality and regression |
| Local-first Apps | Small, useful desktop tools with privacy-first design |
My rule is simple: build for a real problem, make it runnable in five minutes, then improve it through real usage.
One dashboard for your AI coding agents.
Local-first usage aggregation for Claude Code, Codex, OpenCode, Copilot CLI, Qoder, Hermes and more.
Turn requirements into structured test assets with AI.
An AI quality workbench evolving toward RAG, Agent, MCP and regression evaluation.
Turn the workday into a progress bar.
A local-first Tauri desktop app for workday progress, reminders, widgets and Chinese work schedules.
- agent-memory-hub — Share one Git-backed memory across multiple AI agents.
- wise-council-mcp — A multi-perspective MCP decision council.
- animation-memory-museum — A digital childhood museum for Chinese animation memories.
- VoyageAI — AI-powered travel planning.
- project-showcase — Bilingual project showcase site.
Python · TypeScript · React · Next.js · Vue · FastAPI · Tauri · Playwright · pytest · Docker · MCP · RAG
I care more about shipping usable tools than collecting framework logos.
- AI coding observability — usage, cost, sessions and model-level analytics.
- AI quality engineering — evaluation, regression, reliability and test automation for AI systems.
- Small local-first software — simple tools with clear value and low setup cost.
I write about AI engineering, software quality, testing and side projects.
Blog: seanwalter.top/knowledge
Build small. Solve real problems. Keep shipping.