Preview gallery
Similar Skills
View allWhat you get
- One 1024×1536 portrait per request, with tiled grounds, inked and painted objects and a painted animal at the center, never built from tile
- Five dials (palette, indoor or outdoor, animal, mosaic grain, viewpoint) while the rest of the style holds still across a series of pictures
- The eight reference pictures in the package, one sent with each request, and a record of the dial values used for each result
Good fit
- Article headers, newsletters and homely lifestyle posts needing one consistent series look
- A gift picture of someone's pet, set on a slow sunny afternoon at the center of the scene
- Editors and writers happy to hand over one line and receive finished art in return
Not for
- Anyone with neither a FAL account nor a Codex or ChatGPT image tool: no picture can be made
- Other looks, such as photorealism, flat vector art or an animal assembled from tiles
- Landscape-first layouts, since the standard frame is a tall 1024×1536 portrait
About this Skill

One line about a place, some things and an animal is all this Fable 5 skill asks for. It grows that into a single specific moment, sets five dials, and paints it in a locked look: a ground or wall of small uneven tiles, 5 to 9 objects loose on one surface in paint with firm ink edges, a patterned cloth somewhere in frame, and an animal resting where the feeling gathers. The picture comes from gpt-image-2, image-to-image from one of eight bundled references, as a 1024×1536 portrait.
How it works
- Expand your few words into one rich, exact moment, with no questions back.
- Choose a palette, setting, animal, grain and a 3/4 or overhead view.
- Lay 5 to 9 objects on one surface, show a patterned cloth, and give the animal the center.
- Make the picture image-to-image from one reference at high fidelity and high quality in portrait.
- Use the correction guide on faint tiles, a tiled animal, clutter or lines that look too clean.
Checked before delivery
- At least one large ground or backdrop in small, unevenly toned square tiles; objects and animal painted, never tiled
- 5-9 objects loosely placed with air around them, a patterned textile showing and the animal at the emotional center
- A calm, nostalgic, rural mood with at least one cream or neutral tone in the palette
- No words beyond tiny handwritten-looking object labels, and nothing taken over from the reference pictures
What you need before you start
- You bring
- A scene brief naming the place, a mood or some objects, and which animal belongs at its heart
- Dependencies
- FAL_KEY for gpt-image-2, or the image tool built into Codex or ChatGPT
- Outside costs
- FAL_KEY
- Last tested
- v1.0.0 — Claude Code / Codex / Cursor
User reviews
Loading reviews…
FAQ
Is a FAL key mandatory?
Only outside Codex and ChatGPT. If your host has its own image tool and you choose it, no key or paid task is involved. In any other case the picture comes from gpt-image-2 on fal.ai with your FAL_KEY.
How much do I need to write?
A line is plenty, such as "foggy morning, plums and an old spaniel". Your agent builds the full moment from it and does not ask follow-up questions.
Will a set of pictures look related?
Yes. Only the dials and props change from picture to picture, and each request starts from one of the eight references, so tiles, paintwork and ink stay consistent.
Which settings are mine to change?
Palette, indoor or outdoor setting, animal, mosaic grain from fine to chunky, and viewpoint, either a 3/4 angle or straight down. The rest of the style is locked.









