quality-integration-test

Multi-layer integration testing is automated through a CSV-wave orchestration pipeline that validates codebases across L0 to L3. It applies a deterministic phase-based approach with adaptive strategy looping to converge on required coverage targets.

CSV-wave orchestration generates per-wave task files and executes them via agent spawning to validate codebases iteratively. Findings and context propagate from each wave into the next, enabling adaptive strategy adjustments until coverage converges.

Self-iterating integration tests run in auto mode with configurable concurrency, iterations, and coverage targets. The pipeline generates CSV tasks per wave, orchestrates agents, and aggregates results into master tasks for continuous improvement.

Adaptive strategy looping validates a new integration path from code discovery through E2E verification with reflective iteration. Discoveries from each wave propagate context forward, adjusting the strategy until coverage targets are met.

Multi-agent orchestration propagates context from findings across each CSV wave. Agents spawned per-wave execute tasks, and results aggregate into master tasks, feeding discoveries into subsequent waves for continuous strategy refinement.

Phased validation pipelines require codebases suited for L0-L3 layer validation and depend on deterministic phase-based execution. Auto mode assumes default targets are acceptable, and convergence depends on the adaptive strategy's ability to reach coverage goals.