← All learners
06 / The proof · Developers
Maya R.
Mid-career backend developer (Python/Go), new to Claude Code, migrating her team's brittle CI review pipeline into an agent-assisted workflow.
Learning goal
“Run one real PR through an agent-assisted review I actually trust.”
Profile
Cell: Claude Code × AI agents
Elected referents: NY Mets · Fania-era salsa
Accommodation: info-at-point-of-use
Starting mastery: AIHC.2.A1 (0.55) · AIHC.2.B1 (0.40) · AIHC.1.D1 (0.30)
● Docent note — what this arc demonstrates
Twelve lessons: one diagnostic, then one per AIHC anchor from Specification through Regeneration. Every lesson is generated on Maya R.'s own real work — swap in another learner and the Do's and exit tickets change materially. Read straight through, or jump to any anchor. The arc ends in a learning ledger that tracks mastery by standard.
01
Diagnostic — Her CI Migration As Baseline
02
A1 Specification — Contract Specs For The Fee-Rounding Fix
03
A2 Delegation — Sorting The Review Checklist By Reversibility
04
A3 Elicitation — Asking The Reviewer What It's Unsure Of
05
B1 Verification — Running The Contract's Own Checks On The Fee Fix
06
B2 Failure Literacy — A Failure Log For The Agent-Assisted Pipeline
07
B3 Trust Calibration — Permission Lanes, Not Vibes, For payments-service
08
C1 Provenance — Who Wrote This Diff: Attribution For Agent-Produced Code
09
C2 Governance — Hooks Enforce What The Checklist Only Advised
10
C3 Measurement — Defect Catch-Rate, Not Completion
11
D1 Orchestration — Splitting The CI Migration Across Subagents
12
D2 Regeneration — The Update Drill: What Re-Verifies When Claude Code Ships