verify-work

Conversational testing for feature verification works by guiding you through test scenarios, presenting expected outcomes, and capturing your feedback to document deviations in a persistent UAT.md file. You initiate a session and the system extracts deliverables from SUMMARY.md to begin.

UAT tracking records test progress in a persistent UAT.md file that survives session resets. It maintains state across testing phases by logging expected outcomes, user-confirmed deviations, and automatically inferred issue severities for future reference.

You can validate built features through interactive conversational testing. The system presents expected behaviors for each deliverable, processes your verbal or text feedback on whether the feature matches expectations, and automatically logs discrepancies with inferred severity levels.

Yes, conversational AI automates UAT documentation by driving the testing session and creating a UAT.md file. It extracts testable deliverables, presents them sequentially, records your validation responses, and feeds identified gaps directly into the planning phase.

Issue severity inference during manual QA testing automatically evaluates user-reported deviations against expected outcomes. When you describe a discrepancy during the conversational test, the system infers the severity level and logs it into the UAT.md file for further action.

Conversational feature verification relies on human input to confirm deviations and requires a SUMMARY.md file to extract testable deliverables. It does not execute automated test scripts but instead structures manual validation into a trackable UAT.md format.