Companion repository for the Code with Antonio video course · Watch on YouTube.
Build a multi-agent AI team in Python from scratch: planner, parallel researchers, writer and reviewer. No frameworks.
A research team of AI agents: give it a question, a planner splits it into tasks, three researchers search the web at the same time, a writer drafts a report with sources and a reviewer sends it back until it's approved; finally, a lead agent runs the whole team by itself.
| # | Lesson | Video | Folder |
|---|---|---|---|
| 1 | Why a team of agents? | 0:00 | lessons/01-why-a-team-of-agents |
| 2 | One agent, as a class | 3:02 | lessons/02-one-agent-as-a-class |
| 3 | The planner | 5:59 | lessons/03-the-planner |
| 4 | Workers in parallel | 8:03 | lessons/04-workers-in-parallel |
| 5 | Writer and reviewer | 10:39 | lessons/05-writer-and-reviewer |
| 6 | The lead: agents as tools | 13:58 | lessons/06-the-lead-agents-as-tools |
| 7 | Is it worth it? | 16:26 | lessons/07-is-it-worth-it |
- OpenAI Python SDK 3.24.0 · Responses API · gpt-6.1-sol + gpt-6-luna · web_search · pydantic 2.13.5 · uv 0.12.14 · Python 3.12 · Ubuntu (Linux)
- Python 3.12+ and uv
cd app
uv sync
cp .env.example .env # then put your OpenAI API key in .env
uv run --env-file .env solo.py "FastAPI, Django or Flask for a beginner's first API in 2026?" # one agent
uv run --env-file .env team.py "FastAPI, Django or Flask for a beginner's first API in 2026?" # the team (planner, researchers, writer, reviewer)
uv run --env-file .env lead.py "SQLite or PostgreSQL for a small side project in 2026?" # a lead agent that calls the team as tools
# the notes and the report land in workspace/README.md this file
lessons/ one folder per lesson: goal, timestamp, commands and files exactly as in the video
app/ the final working project
assets/ thumbnail
MIT — use it for anything, credit appreciated.
Generated on 2026-10-05 from the course scripts. Code with Antonio · AI tutor.