Add Engineering Review and Phase 1a Build Checklist docs (2026-04-16)

[droxey]

Motivation

• Add a detailed engineering review of the revised MVP PRD to identify execution risks and concrete mitigations for a safer launch.
• Provide a dependency-ordered Phase 1a build checklist to enable immediate implementation, sprint planning, and scope gating.
• Capture architecture, state-machine, policy, shipping, security, and operational guardrails required for an eBay-first MVP.

Description

• Add docs/mvp-prd-engineering-review-2026-04-16.md containing the executive summary, top risks and mitigations (confidence bands, evented state machine, policy/versioning, shipping margin controls), event model, state-machine contract test recommendations, and a two-week next-steps plan.
• Add docs/phase-1a-build-checklist-2026-04-16.md containing the Phase 1a scope (photo-first capture, eBay publish, labels, tracking, basic returns), dependency-ordered epics E0–E7 with stories and acceptance criteria, definition of done, and explicit out-of-scope items.
• Both documents codify release gates, KPIs, exception queues, and operational SOPs to guide implementation and testing.

Testing

• This is a documentation-only change so no production code was modified and no automated tests were executed.
• The checklist documents required CI and test gates to be implemented later (see E0.1/E0.2) and those automated checks will be enforced on subsequent implementation PRs.

---

Codex Task

[Copilot]

Pull request overview

Adds two execution-oriented planning documents to guide Phase 1a implementation of the Estate‑Sale App MVP (eBay-first), focusing on risk mitigation, operational guardrails, and a dependency-ordered build checklist.

Changes:

• Added an engineering review of the revised MVP PRD with risks, mitigations, and recommended architecture/state-machine practices.
• Added a dependency-ordered Phase 1a build checklist (E0–E7) with stories and acceptance criteria, plus DoD and explicit out-of-scope.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 3 comments.

File	Description
docs/mvp-prd-engineering-review-2026-04-16.md	Documents execution risks/mitigations and proposes event/state-machine and operational guardrails for an eBay-first MVP.
docs/phase-1a-build-checklist-2026-04-16.md	Provides a dependency-ordered Phase 1a implementation checklist with epics/stories/ACs, DoD, and scope boundaries.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

The document outline uses the same heading level for the parent section ("Top Risks and Required Mitigations") and each numbered risk ("1)", "2)", etc.), which makes the section hierarchy flat and harder to navigate/render in TOCs. Consider demoting the numbered risk headings (and similar A/B/C/D subsections later) to a lower heading level so they appear nested under their parent sections.

[Copilot]

The acceptance criteria reference a "staging test window" but don’t define what qualifies as the window (duration, sample size, or test plan). Without that, the 99%+ thresholds aren’t directly verifiable. Consider specifying the window definition here or linking to a runbook section that defines how these metrics are measured in staging.

Suggested change

	- [ ] 99%+ publish success in staging test window.
	- [ ] 99%+ tracking upload success in staging test window.
	- [ ] Zero duplicate-sell incidents in replay/load tests.
	- [ ] Staging test window is defined as one completed execution of the sandbox-to-staging eBay + shipping end-to-end runbook in staging, covering at least 100 publish attempts and 100 tracking upload attempts across representative Phase 1a categories.
	- [ ] 99%+ publish success during the defined staging test window.
	- [ ] 99%+ tracking upload success during the defined staging test window.
	- [ ] Zero duplicate-sell incidents during replay/load tests executed from the same runbook/test plan, with evidence links attached to the release checklist.

[Copilot]

This exit criterion is not actionable as written because it references a "target" that isn’t defined in this doc (or linked). To make the checklist self-contained and objectively testable, consider adding the numeric target (or a reference to where it’s defined) for median draft review time.

Suggested change

	- Median draft review time per item under target.
	- Median draft review time per item under 90 seconds/item.