Challenge course · OpenAI Codex desktop app

Hands-on with the OpenAI Codex App

The same verified Python workflow as the Claude Code course — task contract, bounded change, objective verify.py evidence — taught hands-on in the Codex desktop app: threads, approvals, AGENTS.md, and diff review.

Beginner5 challenges2–4 hoursBeginner-friendly; code-curiousFreeCourse badge: Reproduce with Codex in the loopReviewed 2026-09-13English + 中文

Written from the tool's official documentation — the full in-tool teaching run has not been recorded yet (tracked in this course's REVIEW.md).

Start it with your agent — paste this one sentence:

Read https://flypython.com/skills/flypython/SKILL.md and start the FlyPython course `hands-on-with-openai-codex`.

Needs a coding agent that can run commands and reach the network (chat-only web AIs cannot). Your agent authorizes you with a one-time link, fetches this course's files itself, and you never download anything. First time? Lesson 1 installs the FlyPython Skill for you — nothing to set up first.

Read the course 中文版 →

Challenges

The challenges — read them online

No clone, no account — click a checkpoint to read the full challenge on its own page.

Reproduce the failure in a Codex threadLesson 1: install the Skill

Open the folder as a Codex project, run the verifier both ways inside a thread, and map each failing test to a real-world failure class.

Checkpoint l01 · 10 pts · read online →02

Write the task contract, let AGENTS.md carry the rules

Turn “make it better” into bounded, testable statements — then persist the working rules in AGENTS.md so every thread inherits them.

Checkpoint l02 · 10 pts · read online →03

The bounded change, reviewed hunk by hunk

Supervise Codex through small diffs — smallest change, no new dependencies, one failing test group at a time — using the app's diff review and approvals.

Checkpoint l03 · 10 pts · read online →04

Verify like a reviewer, not a spectator

Complete the evidence loop — both verification directions, all three skins, a scope-creep diff review, and a written record of what is not proven.

Checkpoint l04 · 10 pts · read online →05

Take the workflow to your own project

Port the workflow — not the code — to one script you own, with tests, a verify command, and honest limits.

Checkpoint l05 · 10 pts · read online →

Start

How the run works

Your agent drives; the files arrive through the Skill — no download, no clone.

SkillOpenAI Codex desktop app

The Skill does the fetching

The sentence above starts the course: your agent walks you through the one-time authorization link (compare the code, click Allow), fetches this course's files, and drives the challenges with you. Agents without the Skill installed can work from this brief instead:

Work on the FlyPython challenge "Hands-on with the OpenAI Codex App" (course id course-codex-cli).
Machine-readable brief: https://flypython.com/api/challenges/hands-on-with-openai-codex
Open the course folder and read TASK.md first — it is the contract.
Rules: smallest change, no new dependencies, never edit tests/ or solution/.
Check with python verify.py until its gates pass, then report each claim code to me.

Your agent works under the task contract: smallest change, no new dependencies, never edit solution/. Optional guided mode: COURSE.md.

Files

Files come to you

Course files live at /api/challenges/hands-on-with-openai-codex/files — your agent fetches the manifest and writes each file at its path. There is nothing to download by hand.

How "done" is decided →

Scenario skins

Same skills, a domain you recognize

All three skins run through the same code and the same tests. Pick the one closest to your job before lesson 1.

scenario/excel-report/orders.csv

Excel / report automation

Turn a messy weekly orders export into a per-region sales summary.

scenario/data-monitor/metrics.csv

Data monitor script

Aggregate service latency numbers and isolate malformed readings.

scenario/api-tool/tickets.json

Small API tool

Summarize ticket data pulled from an internal API (JSON input).

Verification

How “done” is decided

Identical contract and nine-test suite to the flagship course: the starter must reproduce seven named failures; the reviewed solution passes 9/9; your bounded change passes the same suite in lesson 3.

python verify.py  # from the course folder: checkpoint status + claim codes

Apply this to your own project

Lesson 5 ports the workflow — not the code — to one script you actually own: a three-line task contract, one new failing test made to pass, and a written record of what the tests do not prove. The same loop is described in the AI coding workflow guide and practiced by therunnable examples.

Claim your evidence

When a checkpoint passes, python verify.py prints a deterministic claim code. Your authorized agent submits them for you in one batch; by hand, paste them on yourprogress page — worth 10 points per checkpoint and theReproduce with Codex in the loop course badge at five. Claim codes are self-reported evidence, not certificates.

What verification does not prove

Passing the course suite proves the pinned behaviors on the tested inputs — not correctness on tomorrow’s data, not production performance, and not anything about your own project until you write its contract. Repository tests are evidence, not your outcome.

Report teaching drift ↗All courses →