Studio Trace
(Live product)
“Made with [insert AI tool]”
This label doesn’t tell you anything about the creative’s decision process, judgment or taste.
That is why I made this little side project.
An easy flow to record how AI took part in their work: intent, the role of AI, what they accepted, what they rejected, what they verified, and which decisions stayed theirs.
Live at:
studio-trace.netlify.app
Problem space
Creative work made with AI gets disclosed the way an ingredient gets disclosed on a package.
It doesn’t tell you what the human behind it actually did.
Design decisions
Five questions for the creator, ordered the way judgment happens
The sequence is the product. Intent is captured before AI enters, so the goal cannot be written backwards from whatever the tool produced. Verification comes after judgment, because you can only check a claim you have already committed to.
THE SEQUENCE
- 01
Frame the work
Intent recorded before the tool enters.
- 02
Name the AI role
The tool, the request, and the stage.
- 03
Show your judgment
What was kept and what was refused, together.
- 04
Record verification
Checked against what, and what stays open.
- 05
Claim your authorship
The decisions that stayed with the person.
THE RECORD ASSEMBLES AS YOU ANSWER THE QUESTIONS
- 1 / 5 TRACED
- 2 / 5 TRACED
- 3 / 5 TRACED
- 4 / 5 TRACED
- 5 / 5 TRACED
Each answer lands in the record as it is written. Nothing is written on the creator’s behalf.
Shipped + limits
- SHIPPED
- studio-trace.netlify.app
- Public and working end to end, from first prompt to exported Markdown record.
- ARCHITECTURE
- Nothing is uploaded for project privacy
- Drafts stay in the browser and are never uploaded.
- INTERACTION
- Pixel trail
- A cursor trail that leaves a visible mark as you move through the form, echoing the idea of a trace.
Reflection
Friction is not always a usability failure
In the AGI era, five open text fields are a lot to ask for a user. Every instinct said add generation.
But the effort is what produces the genuine human input trail!