Agentic Coding: Run Coding Agents Like an Operator
How to actually drive coding agents at work: pick a harness, control the context window, write rules and skills that stick, wire hooks and MCP, and run whole tickets through parallel multi-agent workflows.
This course is for developers who already ship code and already use an agent casually — and want the operator-level knowledge behind it. Not prompt tricks: the machinery. What the model actually sees, what you can put in front of it, and how to run many agents at once without losing control of the result.
It is the onboarding I give engineers joining my team, written down.
What you'll be able to do
- Pick a harness and a subscription on purpose, knowing what each one locks you into
- Read your own session: every tool definition, rule file and skill that got loaded
- Cut your context window down to what the model needs, and know when it has gone bad
- Keep one set of agent rules that every harness on your machine reads
- Write skills that the agent loads only when it should
- Move mechanical work (format, lint, checks) off the model and into hooks
- Route every external service through one tool layer instead of a dozen MCP servers
- Run a real ticket end to end as a multi-phase workflow: research, plan, parallel implementation, review, browser QA, pull request
What this course assumes
You write code for a living, you have a terminal you like, and you have at least opened Claude Code or Codex once. Examples lean on Claude Code because that is what I use daily, and they say clearly where a feature is Anthropic-only.
How it works
Every lesson is free and open. Read them in order — each one builds on the context model from the lesson before. Your progress is saved in your browser, so you can stop anywhere and pick it up later. No account needed.
Lessons
- 01
Choose Your Coding Agent
4 minPick a harness by working backwards from the subscription that is allowed to drive it, and know which features you lose when you switch.
- 02
How the Model Actually Works
4 minThe model is stateless and emits only text: see the messages array your harness resends every turn, and how to budget a plan with a weekly cap.
- 03
Rules Files and Scope
5 minWrite your agent rules once in ~/.agents/AGENTS.md, symlink every harness to it, and know exactly which files each harness loads and in what order.
- 04
Own Your Context Window
6 minRead what your agent actually loads, cut the tool definitions and bundled extras you never use, and start fresh before quality drops.
- 05
Skills: Teach the Agent Your Way of Working
4 minPackage your repeated instructions as skills: the SKILL.md format, progressive disclosure, prompt vs router vs script-backed shapes, and how to install them safely.
- 06
Hooks: Move Work Off the Model
4 minWire format, lint and guardrail commands into agent lifecycle events so the machine does them every turn instead of the model remembering to.
- 07
MCP Without Tool Bloat
4 minConnect Jira, Notion, PostHog, Stripe and the rest through one gateway, so hundreds of endpoints cost a handful of tool definitions.
- 08
Voice Dictation
4 minDictate prompts instead of typing them: a local speech-to-text app for free, Groq for cents, and one rules line that makes the agent forgive homophones.
- 09
Dynamic Workflows
4 minType ultracode and the model writes a script that spawns sub-agents: phases in sequence, agents in parallel, parent context window almost empty.
- 10
Workflows in Action
5 minOne real ticket end to end: a research prompt that produces a reviewable plan, an execution prompt that ships it, and the diff review that catches what the summary hides.