Creator & Maintainer↗
01
CoreCoder
Built a runnable 1.4K-line Python coding-agent harness.
PythonRunnable agent harness
1.6K+GitHub stars
Yufeng He · 何宇峰
Building reliableagent systemsfrom post-training to inference.
Former Moonshot AI (Kimi) researcher and Mooncake OSS Maintainer, working across coding-agent post-training, agent harnesses, multi-agent systems, and LLM inference.
Agent harnessesCoding-agent post-trainingMulti-agent systemsLLM systems
M.Sc. Computer Science · The University of Hong Kong
Agent systemsOpen sourceHong Kong
370+upstream PRs
200+agent-harness PRs
56inference-system fixes & improvements
30K+stars on Vibe Trading
02 / Open source
Systems I helped
bring to life.
Compact, inspectable systems that turn agent ideas into reliable software— from a runnable coding-agent harness to production-facing finance tooling and inference infrastructure.
Co-initiator & Core contributor↗
02
Vibe Trading
An open-source AI trading research workspace that turns natural-language finance questions into runnable market analysis, strategies, backtests, and reports.
Agentic financeTrading research
30K+GitHub stars
03
Open source at scale
Open-source contributions
across agent and inference systems.
370+upstream PRs across 40+ repositories
Agent harnesses
200+ upstream PRsQwen Code ↗50+ upstream PRsCoding agent
OpenClaw ↗6+ upstream PRsAgent harness
DeerFlow ↗15+ upstream PRsAgent harness
Inference systems
Mooncake OSS MaintainervLLM ↗10+ upstream PRsServing & runtime
Mooncake ↗35+ upstream PRsKV-cache infrastructure
PyTorch ↗8+ upstream PRsCore systems
Triton ↗System contributionsGPU programming
01 / Experience
Research depth.
Engineering range.
Across agent harnesses, coding-agent post-training, multi-agent systems, agentic finance, and LLM infrastructure.
AI Agent Researcher
Agent research that survives contact with engineering.
- Researched agent harnesses and runtimes at Kimi Alpha Lab.
- Contributed to Kimi Coding Agent post-training; independently built long-horizon coding benchmarks.
- Developed multi-agent and agent-swarm methods; ran extensive evolutionary experiments.
- Core contributor to a confidential multi-agent quantitative-finance project; led orchestration, trajectory evaluation, and long-horizon reliability.
01
Baidu
AI Agent Engineer
AI Talent Program
Co-developed an ad-bidding agent that raised ROI target attainment from 81% to 101%.
02
Maimai
AI Agent Engineer
CEO Mentorship
Independently built a skill-tagging agent under CEO mentorship.
03
Kuaishou
AI Agent Engineer
Transcoding scheduling
Built an ROI-aware transcoding scheduler for 100K+ CPU cores.
05 / Research
Studying how
agent systems fail.
My research focuses on evaluation blind spots: what long trajectories hide, how multi-agent systems break, and where benchmark scores stop telling the truth.
R/01
Sole author
TrajBias ↗
Built a long-horizon coding-agent trajectory-evaluation framework and identified recovery, temporal, and outcome biases.
R/02
First author
FinAgentFail
Built a financial multi-agent failure benchmark with a nine-mode taxonomy.
R/03
Sole author
FMGAN ↗
Adapted R3GAN to time-series imputation.
04 / Selected projects
Built beyond
the benchmark.
Two earlier systems that stretched from edge AI and digital twins to multi-agent music generation—and taught me how to own an engineering outcome.
CTO · May 2025 — Jul 202601
Edge AI · Health screening
FormulaSperm
Built an edge-AI health-screening system and achieved sub-3% digital-twin simulation error.
Co-Founder & Full-Stack Engineer
Oct 2021 — Jun 202202
Oct 2021 — Jun 202202
Generative AI · Music systems
Qiyin
Built a multi-agent music engine that won National Gold at China's “Internet+” competition.
03 / Education
Sep 2025 — Nov 2026
The University of Hong Kong
M.Sc. in Computer Science
Sep 2021 — Jun 2025
Beijing University of Posts and Telecommunications
B.Eng. in Telecommunications Engineering
Selected honors
- Champion2026 HKU DeepTech100 · HK$1M+ prize
- Captain · 1st Place (Gold)BUPT · 2023 Shanghai Global AI Innovation Competition · LLM Track
- 3× SilverICPC Asia Regional Contests
- National First PrizeNational Olympiad in Informatics in Provinces (NOIP)
03B / Skills & interests
Fluent across the stack.
Languages
Python / C++ (primary) · Java / Go
Agent systems
Agent harnesses · Coding-agent post-training · LLM systems
Practice
Claude Code / Codex power user · English (professional working proficiency)
Interest
Top-10 finisher · China Youth Go Open