Write a Critique Brief Before Asking for Feedback
The AZET team is the team writing for AZET, the operator of azet.io. This illustrative guide is for reference and creative exploration, not a demonstrated AZET feature; all creative works, reviewer comments, and exchanges below are hypothetical.
As of 2026-10-12
“Tell me what you think” accompanies a six-panel comic. One reviewer rewrites the dialogue. Another suggests new colors. The maker only wanted to know whether the last panel made sense. All three people did work, but they worked on different questions.
We would send a critique brief with the comic. Its purpose is to give human reviewers a clear reading task and give the maker a usable way to receive the answer.
Name the decision this round can change
In this hypothetical comic, a gardener leaves a cup on a wall. A bird nudges it toward a waiting plant. The final panel reveals that the cup is empty. The maker wants the reveal to feel gently comic rather than confusing.
At this stage, panel order and speech bubbles can change. The page size and monochrome treatment are settled for this draft. Saying that upfront helps reviewers spend their attention where a revision is possible.
A scope is a request, not a muzzle. Leave room for a brief out-of-scope observation if something materially interferes with reading. Ask the reviewer to label it separately so it does not swallow the main discussion.
The hypothetical brief that travels with the work
Work: Comic C-02, six-panel comic, draft 2
Reader: Someone who has not seen draft 1
Purpose of this round: Check the sequence and final reveal
Intended tone: Quietly comic, without a spoken explanation
Can change: Panel order, cropping, short dialogue
Keep for this round: Six panels, monochrome, existing characters
First pass:
Read once before opening the maker's explanation.
Describe what you believe happens, in your own words.
Questions:
1. At which panel did you understand the bird's intention?
2. What did you think was inside the cup before the last panel?
3. Where did you pause because the action was unclear?
Not requested:
Color suggestions, a new premise, or a full rewrite.
Response shape:
One reading summary, up to three panel-specific observations,
and optional suggestions clearly separated from observations.
The neutral work ID keeps the title from revealing the ending. The brief does not tell the reader that the cup must seem full until the ending. That would plant the intended reading before the review begins. The maker's explanation can be shared after the first response is captured.
Our editorial preference is to ask what the reader saw before asking what the reader would change. The two answers describe different kinds of information.
Make a comment traceable to the page
Imagine a hypothetical response: “The ending is confusing.” It is a useful signal, but it does not yet identify where revision should start.
A follow-up could ask, “What did you believe happened between panels four and five?” Suppose the reviewer answers, “I thought the gardener had poured the water out off-page.” Now the maker has a specific reading to compare with the drawing.
Separate that observation from a proposed repair. “Show the cup's interior in panel four” is one suggestion. It might clarify the object while revealing the ending too early. The maker can accept the observation and choose another solution, such as changing the hand position in panel three.
Keep a small response note with three columns: observed reading, possible cause, and revision to try. Mark the cause as a hypothesis until a revised reading supports it.
Ask AI to sharpen the brief, not impersonate the audience
A bounded drafting prompt might read:
Help edit this critique brief for a hypothetical six-panel comic. The intended reviewers are humans. Preserve the maker's listed constraints. Replace broad questions with questions that invite specific observations about sequence and comprehension. Do not invent reviewer reactions, score the work, or rewrite the comic. Flag questions that reveal the intended ending before the first read.
Use the returned questions as candidates. A smooth question can still lead the witness: “How effectively does the final panel reveal the empty cup?” assumes both that the reader noticed the reveal and that it worked.
“Describe what changes in the final panel” leaves more room for an unexpected reading. You can compare that response with your intention afterward.
Review the revision without coaching
The common failure is arguing the reviewer into understanding the original draft. Once you explain the action, they are no longer reading it with fresh eyes. Their later agreement cannot tell you whether the revised panels communicate on their own.
Save the first response, make one focused revision, and ask a new reader to use the same brief. Keep the draft number attached to each response. Otherwise comments about the older cup angle may get mixed with comments about the new one.
If reviewers disagree, preserve the disagreement. Look for the visual or textual cue each person used. You do not need to average their taste into a single verdict or accept every suggested change.
FAQ
How many questions belong in a brief?
For this small comic, three focused questions are enough to start. A larger work may need separate review rounds. Prefer questions you can act on during the current revision.
Can I ask reviewers to avoid harsh language?
Yes. Request comments about the work and specific reading effects. You can invite candor while asking people to avoid personal judgments about the maker.
What if a reviewer ignores the scope?
Thank them for the contribution and identify the part relevant to this round. Keep other suggestions separately if useful. You remain responsible for choosing which revisions serve the work.
Our maker introduction dated October 3 records an email waitlist for the AI assistant. Visit azet.io for the waitlist and current notices if you want to follow the project. This creative exercise is not a live availability check for October 12.