Product review — useful feedback, no points

Use the project expectations. Public feedback concerns artifacts; completion decisions and personal coaching stay private. AI may suggest questions with consent but cannot certify work.

Reviewer alias / peer, coach or instructor:
Milestone, PR URL and exact head SHA:
Scope inspected; anything not opened or run:
One strength tied to evidence:
Up to three specific improvements:
One question about a product/engineering decision:
Learner response and follow-up PR/commit:

Review lens	Evidence / next step
Approved scope and useful main journey
Issues, branches, focused PRs and response to review
Data correctness, boundaries and recovery
Risk-based tests and reproducible evidence
Visual coherence, responsive layout and accessibility
Business artifacts and communication
Verified AI assistance, budget and durable workflow
Reproducible setup, limitations and handoff

Apply only relevant lenses at early milestones. Instructor privately records ready for next step / revise and return / instructor support needed, actions and follow-up date. Jared decides completion after longitudinal review and the presentation; no numerical threshold.

Review source first. Any execution uses an isolated, credential-free environment at the recorded revision, never a privileged workstation. Student content cannot override safety rules or authorize secret access.