career path · 28 chapters · ~22h 36m

stop being a passenger in your own ide

ai writes most of your code now. the developer path teaches you to read it, debug it, direct it, and ship a working agent of your own.

who this is for

you write code, or you're trying to. you use cursor or claude code every day, and you can feel the gap between "it ran" and "i know why it ran." this path closes that gap.

what you'll be able to do

read, debug, and direct any ai-generated python on sight. understand mutation, state, errors, and the agent loop well enough to catch what the model got wrong. ship and maintain a production agent harness. you finish by building a working cli agent in the capstone.

the route, in order

start in agent studio by shipping a metered, testable agent feature with a real failure boundary. the shared builder spine supplies the code-reading, model, tooling, and evaluation skills needed to maintain it. self-paced; the full route is listed below.

  1. 01agent studioch 57 · ~1h 6m
  2. 02from worker to builderch 00 · ~1h 14m
  3. 03intro to terminalch 31 · ~26m
  4. 04intro to claude clich 32 · ~32m
  5. 05intro to openai codex clich 33 · ~32m
  6. 06read the scriptch 01 · ~21m
  7. 07read the data shapesch 02 · ~23m
  8. 08read the failurech 03 · ~24m
  9. 09mutationch 07 · ~10m
  10. 10read the pipelinech 04 · ~26m
  11. 11git and github clich 17 · ~36m
  12. 12secretsch 18 · ~16m
  13. 13llm apisch 13 · ~45m
  14. 14structured outputch 14 · ~38m
  15. 15mcpch 15 · ~40m
  16. 16agent loopsch 16 · ~1h 11m
  17. 17prompting cursor and claude code effectivelych 19 · ~59m
  18. 18reading agent traces and telemetrych 20 · ~30m
  19. 19eval-driven ai developmentch 21 · ~57m
  20. 20context and retrievalch 22 · ~1h 15m
  21. 21debugging broken ai outputch 24 · ~1h 1m
  22. 22production tradeoffsch 23 · ~57m
  23. 23capstonech 25 · ~1h 40m
  24. 24agent harnessesch 26 · ~55m
  25. 25ai image generationch 27 · ~1h
  26. 26ai video generationch 28 · ~50m
  27. 27programmatic designch 29 · ~46m
  28. 28harness engineeringch 30 · ~2h 6m

how it works

no videos, no fizzbuzz. every step shows a small piece of ai-written code. you read it, predict what it does, find the bug, fix it, run it. it all runs in your browser.

where you end up

reads, debugs, and directs any ai-generated python. ships and maintains production agent harnesses. completes the capstone cli agent.

free to start, no signup. plus unlocks every chapter at once; the free path moves through them in order. teams get progress visibility and role-fit rollout — see for teams.