Evidence first.
Scores second.

What happens when you run a review?

1. Choose a project already submitted to the class board. We use its saved GitHub and website links.
2. Read selected public source files: HTML, styles, components, dependencies, tests, README and assignment evidence. File and time limits keep reviews bounded; this is not an exhaustive code audit.
3. Request a technical website scan from LEGIT Agentic. If it is unavailable, the report says so and continues with the source evidence.
4. Include public course quest evidence (not the private university portfolio), clearly labelled as pending, approved or needing revision.
5. Run Utility and Design independently. Save both reports, with a timestamp, inspected files, ideas and coding-agent prompts.

The agents do not execute your application, view screenshots, access your private dashboards, or complete a purchase. A dependency or HTML signal is not proof that a feature works. Inaccessible evidence means “unverified”, not an automatic zero.

Utility: formative feedback on your work

Useful outcome and CTA: 25 points. Working implementation: 35. Tests and evidence: 25. Meaningful improvements: 15. The total is formative feedback out of 100; it is not converted into a university mark out of 10. If a criterion cannot be assessed, the total stays pending.

Use it to improve your work and document your response. Your individual evidence portfolio, including what existed before the course and what you personally changed and tested, supports instructor assessment. It does not set an official university mark or decide whether you pass. The four university component requirements still apply: 15% participation, 30% Digital Presence, 20% Digital Advertising and 35% Experimentation. Submit your individual portfolio to the university’s Connect platform; this checker does not upload or read that private submission. Commit volume and tool count do not prove quality.

Design: feedback, never a pass/fail factor

Hierarchy and clarity: 35 points. Visual consistency: 30. Responsive presentation: 35. The design agent reads source-level styling and layout intent; it explicitly calls out the need for a visual check. Its score never enters the Utility score, university grade or passing decision. Use it to explore ideas, not to conform to one person’s taste.

Sources and additional audits

Checked via public GitHub source, submitted course evidence and LEGIT Agentic when available. Each report names the sources actually used. We do not claim “100+ trusted sources”.

GitHub

Connected source inspection and course commit tracking.

LEGIT Agentic

Connected technical website scan. Availability and results are shown per review.

Open checker ↗

40rty · AgentIQ

Optional external audit. Results are not imported automatically.

Open external audit ↗

Boring Marketing

Optional external AI visibility audit. Results are not imported automatically.

Open external audit ↗

External audit scores are different measurements; we do not average them into a course grade.

Review, improve, repeat

Run from the browser used to register your project, or connect an instructor session. Wait at least two minutes between runs. Push changes before running again; a saved report describes the source inspected at that time. If one agent fails, the other report can still be shown. Your source and previous report are preserved if both fail.

Reports and quest evidence are shared with the class. Include public project material only. Discuss incorrect findings with the instructor and show the relevant code or test evidence.