I build full-stack tools, study AI agents, and write notes so future-me has fewer mysteries to solve.
Most days I am somewhere between a product idea, a half-finished experiment, and a README that still needs one more example.
$ whoami
QiQiLu
$ focus
Full-stack development | AI agents | knowledge graphs
$ current_mood
curious, caffeinated, occasionally overengineering
$ tiny_rule
If I have to explain it twice, it probably needs a README.- Small tools that turn vague ideas into something clickable.
- AI workflows that are debuggable instead of just impressive in demos.
- Knowledge systems where notes, people, projects, and context can actually find each other.
- Interfaces that feel calm, quick, and hard to accidentally break.
| Project | Why it exists | Stack |
|---|---|---|
| ai_study | My AI learning notes, experiments, and "wait, why does this work?" moments. | Python, Jupyter, OpenAI |
| ai-office-graph | An attempt to make office knowledge less scattered and more searchable. | Python, TypeScript, Neo4j |
frontend JavaScript, TypeScript, React, Vue, Next.js, Tailwind CSS
backend Node.js, Python, Java, Express
ai / data OpenAI, LangChain, PyTorch, Jupyter, Neo4j
tooling Docker, Git, GitHub Actions, Linux, VS Code- Write the boring docs before the clever abstraction.
- Name things like someone tired will read them at 1 AM.
- Ship the first version before the perfect version starts giving opinions.
- A good bug report is a tiny detective story with timestamps.
This profile has a small scheduled job that redraws the contribution snake every few hours.
It is not exactly real-time, because GitHub README files cannot run custom JavaScript, but it is alive enough to make the profile feel less like a poster and more like a little status board.
|
GitHub stats, because charts are still nice
I am happy to talk about AI agents, full-stack projects, knowledge tools, or weird bugs that were obvious only after three hours.
- GitHub: @zqzlq
- Email: wy3393675686@163.com