| 1. Creative Vision |
Specific, intentional creative goal before AI; final work reflects it |
Vision present, partially developed; final work mostly reflects it |
Stage 1 brief incomplete or vague |
No evidence of Stage 1 thinking |
| 2. Prompt Quality |
Multi-layer prompt, all layers/elements present, clearly intentional |
Most layers/elements present; minor gaps |
Only 1–2 layers/elements; output generic |
One-sentence prompt; no creative direction |
| 3. AI Collaboration |
2+ iterations, compared, gaps clearly documented |
Iterated at least once; gap analysis surface-level |
Single output, no iteration or comparison |
No evidence of AI Collaboration stage |
| 4. Revision Depth |
5+ revisions across all 3 Layers, each justified |
4–5 revisions, most layers represented |
2–3 revisions, only 1–2 layers, mostly cosmetic |
Fewer than 2 revisions; final barely different from AI draft |
| 5. Human Signature |
Distinctive creative identity evident throughout |
Some elements of distinctive voice visible |
Final work resembles the AI draft closely |
Virtually indistinguishable from the raw AI draft |
| 6. AI Disclosure |
All 6 sections completed, specific and honest |
5 of 6 sections completed; one vague |
3–4 sections completed; human contribution generic |
Absent or only 1–2 sections; no authorship claim |
| 7. Authorship Reflection |
Specific creative insight; names what AI cannot do here |
Present and honest; lacks specificity |
Summarizes what they did, not what they discovered |
Absent or a restatement of the task instructions |