inflearn logo

An AI Agent Team Running on OpenRig: Role-Specific Models (Opus·Haiku·Sonnet·Codex) and Remote Access via Tailscale·RustDesk

If one agent wasn’t enough, bring Claude Code and Codex together as a team. Split up the roles—Opus as lead, Haiku for research, Sonnet for implementation, and Codex for review—to build a small app together. With Tailscale and RustDesk, you can also delegate tasks and check on their progress from your phone or other devices, even when you’re away from your desk.

2 learners are taking this course

Level Basic

Course period Unlimited

AI Agent
AI Agent
codex
codex
LLM
LLM
Network
Network
claude
claude
AI Agent
AI Agent
codex
codex
LLM
LLM
Network
Network
claude
claude
Thumbnail

What you will gain after the course

  • Launch and manage Claude Code and Codex together as one team (rig) with OpenRig.

  • Choosing the right model for each role and the rationale behind it (cost, speed, cross-review)

  • Monitor multiple agents on one screen with Herdr

  • Have the lead take it on once, and have the team split up the investigation, design, implementation, and review.

  • Access my Mac's team from outside without opening ports using Tailscale

  • Safely access your Mac’s screen and transfer files using RustDesk

Recommended for
these people

Who is this course right for?

  • A developer who keeps losing track of who’s doing what while running multiple agents.

  • For those who send every task to the most expensive model and quickly burn through their usage quota

  • For those who want to delegate tasks to an agent and check the results while they’re away.

Need to know before starting?

  • If you’ve ever tried basic terminal commands like `cd` and `ls`, that’s enough.

  • You need a Claude Code or Codex account (subscription and usage costs are your responsibility).

Hello
This is dante

Career Verified

840

Learners

42

Reviews

3

Answers

4.8

Rating

6

Courses

I’ll personally test and teach you how to assign tasks to AI.

Hello, I’m Dante. I previously worked as a developer at Naver and Coupang, and I currently work as a developer in Silicon Valley. Every day, I use AI coding tools such as Claude Code and Codex for real work, measuring what works and what doesn’t with data and incorporating those findings into my courses.

✔ Industry Developer — Former developer at Naver and Coupang, currently a Silicon Valley developer
✔ Author — 『The Complete Guide to Vibe Coding with Codex』, 『Relearning React in Depth: Inside React』
✔ Mentor — Over 400 mentoring sessions (practical AI agent implementation, frontend development and learning consulting)
✔ YouTube “Dante” — A channel that directly tests and compares new AI coding tools

This is how I create my courses
1. I don’t just explain things verbally — I show the input → execution → result exactly as they appear on the actual screen.
2. I measure it myself — Instead of saying “this is better,” I measure and show the time, token usage, and number of bugs. Even results that are unfavorable are shown as they are.
3. I finish with visible results — I create deliverables that work all the way through using real public data.
4. I make all the code public — The practice files and solutions are available in a public GitHub repository.

Courses · Mentoring
· Claude Code in Practice — From Installation to Harnesses — https://inf.run/yKj3F
· An Introduction to Dante’s Complete Codex Guide — https://inf.run/dzVtP
· Building an AI Agent from Scratch — RAG · Multi-Agent Systems · MCP — https://inf.run/jY8Bz
· 1:1 Mentoring — https://mentoring.inflearn.com/mentors/1535

Books
“The Complete Guide to Vibe Coding Codex Today” Kyobo Bookstore · YES24 · Aladin
“Inside React: Learning It in Depth Again” Kyobo Bookstore · YES24 · Aladin

More

Curriculum

All

21 lectures ∙ (1hr 59min)

Published: 
Last updated: 

Reviews

Not enough reviews.
Please write a valuable review that helps everyone!

dante's other courses

Check out other courses by the instructor!

Similar courses

Explore other courses in the same field!

Free