coding Phase
TDD with Coding Agents
You generate code from human-written tests
Level 3: Pilot Mid Risk Evaluate
What It Does
AI generates code from tests created by humans.
Test-driven development: tests → implementation
StrategyRadar Notes
Not as good as specs-driven. Has a higher risk because the AI agent has freedom to write the solution for the tests to pass in many different, potentially inconsistent ways. It's still good when combined with specs-driven.
Premium Packs
Get comprehensive guides, templates, and advanced prompts to implement this capability effectively in your organization.