Lesson 2 — Codex as a Disciplined Teammate

Duration: 120 minutes

Included delivery track: Follow LAB_SETUP.md for the supplied files and local simulation. Required activities use these materials; connected deployments are optional extensions. Instructor debrief: answer key. Real OAuth, cloud releases and payments are simulated or discussed through evidence packets. Deck: 02-codex-as-teammate
Lab: Codebase Detective

Learners will be able to

give Codex a task brief with outcome, context, constraints, and completion bar;
ask Codex to investigate an unfamiliar repository without granting broad authority;
inspect a diff and a test result before accepting a proposed change; and
record useful repository knowledge as a durable playbook instead of relying on chat history.

Instructor preparation

Use the included sandbox, R1 and G102 in fixtures/github-issues.json. Distinguish the handout's staff report from the concierge issue: staff denial is correct. Explain that npm test checks fixture integrity, not release readiness. Scope model context; public answers do not replace learner reproduction.

Agenda

Time	Facilitation
0–10	Arrival challenge: rank three prompts from vague to executable; learners improve the weakest one.
10–25	Demonstrate the task brief. Explain ask/plan first for unfamiliar or consequential work, then bounded implementation.
25–40	Live Codex walkthrough: inspect `AGENTS.md`, map a request path, propose an investigation, and state what it cannot know.
40–50	Guided practice: each pair fills in an AI task brief for the reported send-update behavior.
50–60	Break; coaches verify permissions and tool boundaries.
60–95	Lab: investigate a seeded issue. The driver uses Codex; the verifier checks every repository claim and records sources. Switch roles after the plan.
95–110	Teams review one another’s task brief and reject instructions that are vague, unsafe, or missing a completion test.
110–115	Teach the “proposal → review → apply → verify” loop.
115–120	Exit ticket: write a task brief asking Codex to find—not fix—a suspected regression.

Teaching notes

Use practical repository habits: provide relevant file paths and docs, split work into bounded tasks, keep repository instructions current, and run the project’s verification commands after a change. Modern Codex workflows support durable project context and tool-connected tasks; the course keeps human review and explicit boundaries at the center.

Assessment

Score the AI task brief and evidence log. A proposed code patch alone does not earn credit; the learner must explain which test or observable behavior checks it.