T0–T5 model
Name your starting level. Name the next behavior that proves progress.
At Car-Mart Technology Group: T0 = never used Claude. T1 = tried a few prompts, got generic answers. T2 = use weekly, save good prompts in a Project (api-refactor, q3-dashboard, runbook-update, or your own). T3 = use daily, customize Preferences for tone and output. T4 = built into workflows, multi-step, owns a CLAUDE.md per Project. T5 = agentic, multi-Claude, custom tools. The full long-form read lives at hool.dev/aitraining/t0-t5-model.
The next level is not volume. It is observable behavior that a director can see in shipped work — a saved prompt library, a CLAUDE.md that reflects the team's standards, a runbook diff that took 30 seconds instead of 30 minutes.
Do not overclaim. A T2 with a saved prompt library and a Project is more useful to the team than a T3 who has not set Preferences. Read yourself by what you do, not what you know.
I'm a senior backend engineer on the Car-Mart API team (Python, 8 years, 2 repos in scope). Walk me through the T0-T5 behaviors I should adopt THIS WEEK to move from T2 to T3. For each behavior, give me one concrete example from PR review, runbook generation, and on-call summaries. End with the single highest-leverage change I should make Monday morning.In your lane
Dev lane: T2 saves the PR-review prompt in a Project. T3 sets Preferences for code-style + error-handling. T4 writes a CLAUDE.md for the api-refactor Project with the 3 most-referenced helper modules and the team's PR template. Concrete artifact: open Desktop → Project: api-refactor → new chat → paste PR #1247 diff → 'summarize per PR template.'
Data lane: T2 pastes a SQL query, asks for explanation. T3 sets Preferences for tables-first output and SQL with comments. T4 builds a Project for q3-dashboard with the 2 data dictionaries and the 1 sample stakeholder summary. Concrete artifact: open Desktop → Project: q3-dashboard → 'explain this query, then list the 3 edge cases it does not cover.'
Ops/Security lane: T2 pastes a stack trace, asks for a hypothesis. T3 sets Preferences for SRE-tone and actionable steps. T4 builds a Project for runbook-update with the team's runbook library and a CLAUDE.md that flags any escalation-path or on-call-rotation change. Concrete artifact: open Desktop → Project: runbook-update → drop in the new runbook → 'compare to template, flag deltas.'
Pick your T-level honestly. Write one sentence: 'This week I will…' Save it to your team's Slack #ai-training, your personal note, or the 90-day plan export. Then do the thing. The directors and CTO Josh are in the room to see what your team can ship today.