AI Development Methodology

tinker — cross-epic manual-QA queue

Cross-epic test scenarios that span multiple epics or don’t belong to any single one. Epic-specific scenarios live in each epic’s TEST.md.

Active manual-QA scenarios

ID	Scenario	Affects (epics)	Cadence	Status
QA-01	Fresh-machine install timed end-to-end; target < 60s on macOS / Linux / Windows	E01, E03 (pkging)	Per minor release	not-tested
QA-02	Note round-trip across capture → search → edit (with E02 search shipped)	E01, E02	Per minor release	n/a (E02 not yet active)
QA-03	Calendar-event integration (when E02’s calendar feature ships)	E02	Per minor release	n/a (E02 not yet active)

Acceptance-test inventory (pointers to epic TEST.md files)

Epic	TEST.md	Active scenarios
E01 — CLI foundations	epics/E01-cli-foundations/TEST.md	AT-01 through AT-05
E02 — Search + filters (not yet chartered)	(no TEST.md yet)	—
E03 — Install + packaging (not yet chartered)	(no TEST.md yet)	—

Conventions

• QA IDs use QA-## (cross-epic, monotonic) to distinguish from epic-scoped AT-##.
• Per-epic acceptance tests live in epics/<E##-slug>/TEST.md. This file holds only scenarios that span ≥ 2 epics.
• “Status” mirrors the Test enum.
• When a cross-epic scenario passes, it lands in the affected epics’ regression-scenarios sections + an entry here marks the last-verified date.

How this fits the methodology

This file is the operator-facing queue that pairs with the per-epic TEST.md files (see methodology/03_epics.md “Standard epic-folder structure”). The autonomous loop can pick up automated test work; manual QA stays with the human operator and the cadence is enforced before each milestone declaration per methodology/12_milestone_evaluation.md.

This site is open source. Improve this page.