· Lee Hung-yi-inspired teaching-style posts
English posts rewritten with a roadmap-first, intuition-before-mechanism teaching flow inspired by Lee Hung-yi.
This is a complete style-version collection. These pages were not written, reviewed, or endorsed by Lee Hung-yi. Subscribe to the teaching-style RSS feed.
-
agent-07 · Loop engineering is the harness on a timer — Lee Hung-yi-inspired teaching style
A teaching-style walkthrough of schedule, state, spawning, hard oracles, and the signals an autonomous loop cannot game.
-
agent-06 · Self-improving AI is a spectrum, not a switch — Lee Hung-yi-inspired teaching style
A teaching-style map of pseudo-labels, rewards, AI judges, generated tasks, harness improvement, and the Rubicon.
-
agent-05 · Self-correction has three layers: decoding, workflow, and reasoning — Lee Hung-yi-inspired teaching style
A teaching-style explanation of decoding tricks, grounded workflow feedback, and trained reasoning behavior.
-
agent-04 · Harness engineering is how we actually make agents useful — Lee Hung-yi-inspired teaching style
A teaching-style explanation of tools, permissions, workflows, memory, feedback, and verification loops.
-
agent-03 · What AI agents change about research work — Lee Hung-yi-inspired teaching style
A teaching-style tour of research assistants, idea generation, AI review, verification, and human judgment.
-
agent-02 · Multi-agent systems are topology problems — Lee Hung-yi-inspired teaching style
A teaching-style explanation of collaboration topologies, adversarial games, social agents, and coordination limits.
-
agent-01 · Context engineering is the agent's working memory — Lee Hung-yi-inspired teaching style
A teaching-style explanation of prompt-visible workspace, memory, compression, observations, sub-agents, and skills.
-
agent-00 · The AI agent is a computer system around an LLM processor — Lee Hung-yi-inspired teaching style
A teaching-style architecture map of the LLM processor, context, memory, tools, feedback, harness, and agent topology.
-
My setup for running 10+ CLI coding agents at once — Lee Hung-yi-inspired teaching style
A teaching-style rewrite that reframes 10+ CLI coding agents as an operating-system problem: tmux as workspace, Tailscale as private network, Termius as mobile I/O, and a PWA panel as control room.