const khanh = {
location: "Ho Chi Minh City, Vietnam π»π³",
education: "Information Technology @ HCMUS",
interests: ["AI Agents", "Harness Engineering", "Multi-Agent Systems", "Reading"],
currentFocus: "Building autonomous AI research agents",
philosophy: "Making AI agents more reliable, observable, and production-ready"
};π Currently Working On:
- π¦ Lyra - Lightweight Yielding Reasoning Agent: CLI-native coding agent harness
- π Open-Fang - Autonomous AI research agent for AI/Agents/Harness literature
- βοΈ Agent-Forge - Speculative Multi-Agent Orchestration Platform with MCP & DAG workflows
π Learning & Exploring:
- Multi-agent orchestration patterns
- Harness reliability engineering
- Production-grade AI agent systems
- Evaluation frameworks for AI agents
π¦ LyraLightweight Yielding Reasoning Agent - a general-purpose, CLI-native, open-source coding agent harness Tech: Python, AI Agents, CLI |
π Open-FangAutonomous AI research agent specialized for AI/Agents/Harness Engineering literature Tech: Python, Research Automation |
βοΈ Agent-ForgeSpeculative Multi-Agent Orchestration Platform with MCP, A2A, DAG Workflow, Kafka Tech: Java 21, Kafka, Multi-Agent Systems |
π‘οΈ Aegis-OpsSRE / Ops Agent (MVP) for infrastructure operations Tech: Python, SRE, Operations |
β±οΈ GnomonHarness-aware evaluator + closed evolution loop, exposed as portable Harness IR Tech: Evaluation Framework |
π― PivotProduction-grade AI Agent Reliability Harness - unified evaluation, guardrails, deterministic replay Tech: Reliability Engineering |
I'm always interested in collaborating on AI agent projects, harness engineering, or discussing multi-agent systems. Feel free to reach out!
βοΈ From ndqkhanh
"Building the future of autonomous AI agents, one harness at a time."



