COMPUTER SCIENCE · HONG KONG

Steven Huang.

Building & evaluating AI agents.

I’m a CS master’s student at the University of Hong Kong. I build agent tools, run evaluations, and write down what I learn.

Harnesses / Evaluation / Learning from feedback

Pixel-art character at a desk with books, a lamp, and a computerPixel-art character at a desk with books, a lamp, and a computer
THE WORKBENCHA familiar face from my GitHub.

01 / PROJECTS

Selected work

Tools and experiments I’m working on.

All work
01 / TOOLING · EVALUATIONGitHub

skillfit

Compare Skills, rules, and MCP setups on the same tasks. Keep quality, trigger behavior, and token cost in view.

SAME TASKS · PAIRED RUNS
ABaseline
/
BTreatment
QualityBehaviorToken cost
02 / HARNESS · TOOL USEGitHub

DeviceAgent Arena

A testbed for Chinese tool-use agents, where actions change the next state. Check both the outcome and the path taken.

STATEFUL TASKS · CLOSED LOOP
TaskHarnessTools
observation → next decision
Final stateFull trajectory

02 / OPEN SOURCE

Small fixes, shared upstream.

Selected contributions to tools I use and care about.

Contribution history

03 / NOTES

From the notebook

Reading, references, and things worth keeping.

All notes

SAY HELLO

Get in touch.

For a question about my work, or to exchange ideas.

huangshuaidi666@gmail.com