deep-work-orchestrator

Evidence-driven development is a workflow approach that enforces test-driven progress through sequential phases—Brainstorm, Research, Plan, Implement, and Test—collecting mandatory receipts and health reports to verify each stage of an AI development session.

You can manage multi-phase transitions by initializing a session that automates sequential progression across planning, implementation, and testing phases, applying mandatory reviews and TDD enforcement to govern each lifecycle stage.

No specific dependencies are required to run the core workflow orchestration, though it supports optional integration points for scripts, references, and assets when available to enhance task scaffolding and cross-session coordination.

Yes, session management supports cross-session coordination when available, handling session creation, migrations, and lifecycle management to maintain structured TDD progress across long-running, multi-session AI development tasks.

TDD enforcement works by applying mandatory reviews and phase governance that require receipt-based evidence before transitioning from the implementation phase, ensuring disciplined, test-verified progress throughout the development lifecycle.

The best way is to use an automated phase governance system that sequences the Brainstorm and Research phases before planning, applying optional reviews and approvals to ensure structured, evidence-backed preparation prior to implementation.