feedback-loop

A deterministic bug feedback loop is a repeatable validation signal that tests bug behavior without human intervention. You need it when debugging toil requires consistent, automated hypothesis testing across diagnosing and red-green TDD cycles.

To set up an agent-runnable feedback loop for TDD, select a pattern from the catalog and wire a minimal, deterministic path that runs in the repo's AFK context. This produces verifiable commands and artifacts for repeatable red-green cycles.

Yes, deterministic feedback loops integrate seamlessly with `/diagnose`, `/tdd`, and `/draft-contract` workflows. They provide verifiable inputs and artifacts that drive Phase 1 hypothesis testing and subsequent verification-contract drafting.

Building deterministic feedback loops requires disciplined pattern catalog usage to select the simplest reliable loop aligned with project determinism rules. You also need deterministic inputs and integration with the repo's AFK context to produce verifiable commands.

The best way to ensure repeatable outcomes is applying deterministic feedback loop signals across Phase 1 hypothesis testing and downstream steps. This requires disciplined pattern catalog usage and deterministic inputs integrated with AFK context to produce verifiable artifacts.

Inconsistent bug validation results occur when feedback loops lack deterministic inputs or disciplined pattern catalog usage. Ensure your loop runs in the repo's AFK context with minimal, deterministic paths to produce verifiable commands and repeatable outcomes.