SecBot An authorized security-testing workspace with a NestJS backend, Ink terminal UI, MCP integration, and security-focused tool modules. TypeScript NestJS MCP
|
ExecGo An agent-first execution layer for validated, cancellable, observable, and recoverable actions, with DAG scheduling and pluggable executors. Go Runtime Infrastructure
|
MPH-Agent A domain-specific agent that turns natural-language COMSOL requirements into complete simulation models across geometry, physics, meshing, studies, and solving. Python COMSOL Domain Agent
|
Damn Agent A Chinese learning resource for understanding, building, and evaluating AI Agent systems. Documentation Agent Engineering MDX
|
iammm0 My engineering profile: a collection of open-source work across AI Agent systems, security tooling, execution infrastructure, and applied automation. Open Source AI Agent Engineering Profile
|
|
| Area | What I care about |
|---|---|
| Agent systems | Tool calling, planning loops, context, memory, MCP, and multi-agent coordination |
| Execution | Task contracts, DAG scheduling, retries, timeouts, cancellation, and recovery |
| Safety | Authorization boundaries, sandboxing, resource policy, auditability, and human control |
| Product engineering | Clear interfaces, useful defaults, operational feedback, and documentation that helps people ship |
I started from applied physics and moved into software engineering through scientific computing, full-stack systems, and increasingly systems-oriented AI Agent work. That background still shapes how I build: define the model, make the constraints explicit, and verify what happens in the real system.
中文简介:我从应用物理出发,逐步进入软件工程,当前专注于 AI Agent、安全工具与执行基础设施。比起展示“会什么”,我更关心系统能否被验证、被控制,并真正完成任务。
For collaboration, engineering roles, or open-source work, reach me at wisewater5419@gmail.com or open an issue in one of the projects above.