SkillFrame guide
AI photo critique that tells you what to shoot next
You searched because you want to know if a photo is good and how to fix it. A one-off AI critique can answer that once — but it disappears. SkillFrame ties critique to your frame, scores it on composition, light, and story, and tells you what to shoot tomorrow.
The real problem with one-off critique
Most AI photo critique tools give you vague praise on a single upload. You read it, close the tab, and have no idea whether you are actually improving or what to practice next. Generic feedback reads like a horoscope — flattering, forgettable, untrackable.
Why a one-off AI critique is not enough
One critique on one photo is a snapshot. It does not tell you whether composition, light, or story are trending up across a week of shooting. Random feedback is not deliberate practice — there is no loop, no score history, no next assignment.
- Feedback disappears after you close the tab
- No repeatable yardstick across frames
- Nothing names what to shoot tomorrow
- You cannot prove improvement to yourself or anyone else
What a better critique should do
A useful critique references your actual image, scores it on three dimensions, and closes the loop with a specific next assignment. That is the SkillFrame loop: assignment → shoot → upload → critique → score → repeat.
| Dimension | Critique asks |
|---|---|
| Composition | Does the frame guide the eye on purpose? |
| Light | Is the light shaping the subject intentionally? |
| Story | Does the frame say something at a glance? |
How SkillFrame's critique works
- Daily assignment matched to your genre (portrait, landscape, or street)
- Private upload — your frame, not a stock demo
- AI vision critique tied to the image you shot
- Rubric scores on composition, light, and story (0–10 each)
- Longitudinal skill state so you watch trends, not vanity scores
- Model transparency: Free Taste uses GPT-4o-mini; Sprint and Mastery use GPT-4o
Read more about why deliberate practice beats one-off feedback.
Example critique structure
Example copy — not a real user critique
What worked
Your subject reads clearly against the background: shallow depth and side light separate the face from the bus stop glass without feeling cut out. The catchlight in the near eye gives life; the gaze line leads into the negative space on the right.
What to fix tomorrow
The frame is slightly loose on the left — a brighter patch competes with the face. Tomorrow, shoot the same subject one step closer or crop in-camera so the brightest area supports the face, not the edge of the frame.
| Dimension | Score | Note |
|---|---|---|
| Composition | 7 | Subject placement works; left edge needs discipline |
| Light | 8 | Directional window light; background sacrifice is intentional |
| Story | 6 | Moment is quiet but readable; stronger gesture would lift story |
What to shoot next: Re-shoot the same subject with the same light, but fill the frame so nothing brighter than the face sits in the outer third.
What to upload for your first critique
Start with one recent photo you are unsure about — a frame where you felt something was off but could not name it. Pick the genre you actually shoot: portrait, landscape, or street. Avoid heavily edited composites for the first upload; the critique reads light and composition best on a single capture.
- One JPEG or PNG from your camera or phone
- A frame you shot in the last two weeks
- The genre track that matches the subject (portrait, landscape, or street)
A 3-day teaser: feel the loop before Sprint
Not ready for 14 days? Run three reps to see how critique compounds:
- Day 1: Upload your strongest recent frame as a baseline.
- Day 2: Re-shoot the weakest dimension the critique named (composition, light, or story).
- Day 3: Same subject, opposite light. Compare scores across all three uploads.
Start with a free Taste critique
FAQ
- Is the first critique free?
- Yes. The Taste critique at /try is one free critique with no account required. It uses GPT-4o-mini and scores your frame on composition, light, and story.
- What does SkillFrame score?
- Every critique scores composition, light, and story on a 0–10 scale. Those scores roll into a longitudinal skill state so you can watch trends across uploads.
- Does AI critique replace a human coach?
- No. AI critique is best inside a daily loop with measurable progress. SkillFrame Coaching adds human feedback (video call, portfolio review) inside that same loop when you want it.
- Which genres are supported?
- Portrait, landscape, and street. Each genre has its own assignment wording but the same rubric and skill graph.
- How is this different from uploading to /try directly?
- This page explains the improvement loop behind the tool. /try is where you upload. Start here if you want context; go straight to /try if you are ready to submit a frame now.