Session 3 — AI-Assisted Development Foundations

Phase: 2 — Enablement & Integration Target Week: Week 3 Duration: 2.5 hours Format: Hands-on workshop (developers on their own machines) Audience: Turner development team

Objective

Give every developer fluency with the core AI-assisted development surfaces they'll use day-to-day. Every exercise in this session runs against real code from the learning bed product so that developers leave with both technique and a head start on tasks they actually have on their plate.

Key Topics

Claude product surfaces overview: Claude Code (CLI/IDE), Claude.ai, Claude in Chrome — when to reach for each
Account, workspace, and permissions setup against Turner's environment
Prompt engineering fundamentals: clear goals, providing context, iterating, and verifying
Core developer workflows: explain this code, refactor, generate tests, write documentation, debug a failure
Code review patterns: getting a useful second opinion from Claude on a real PR
Common failure modes and how to recover (hallucinated APIs, stale context, over-eager edits)
When NOT to use AI — knowing the edges of the tool

Outcomes

Every developer has Claude set up and working against the learning bed's repo
Each developer has completed at least three guided exercises against real learning bed code
Shared vocabulary for prompting patterns that will be reinforced in Sessions 4-6
Initial entries captured for the Developer Reference Guide, using learning bed examples

Inputs Required from Turner

Developers on the call with their actual dev environment available
Local clone of the learning bed product's repo on each developer's machine
Any required Claude licensing/access provisioned before the session

Deliverable Contribution

Anchors the Developer Reference Guide (Phase 2