Session 6 — Speed vs. Risk: Capstone & Best Practices

Phase: 2 — Enablement & Integration Target Week: Week 5 Duration: 2.5 hours Format: Capstone exercise plus best-practices discussion Audience: Turner development team, technical leadership (recommended)

Objective

Operationalize the velocity-vs-quality discussion on the learning bed product: developers take a real backlog item from the learning bed end-to-end with AI assistance under Turner's governance framework, then close the engagement with a synthesis of what's working and a plan for replicating the wrapping across the rest of the portfolio.

By the end of this session, the learning bed product is fully wrapped in AI-assisted development practices — CLAUDE.md and skills in the repo, AI-augmented pipeline running, a real backlog item completed against the new toolchain. The team now has the worked example they need to do the same on every other product they own.

Key Topics

  • Velocity-vs-quality patterns: shipping fast without compounding technical debt
  • Security guardrails in AI-assisted workflows: secret hygiene, dependency vetting, supply-chain awareness
  • Code review etiquette when AI is in the loop — human and AI reviewers complementing each other
  • Capstone exercise: take a real backlog item from the learning bed from refinement through PR using the full toolchain (CLAUDE.md, skills, pipeline integration)
  • Retrospective: what worked, what was friction, where Phase 3 support should focus
  • Hand-off of the finalized Developer Reference Guide
  • Forward-looking roadmap: which products Turner should wrap next, in what order, using the learning bed as the template

Outcomes

  • A real learning bed backlog item completed during the session with full AI-assisted workflow
  • Finalized Developer Reference Guide ready for hand-off
  • Pr