Childhood Me example output generated by Doodle AI — A real run of this skill. The input is a synthetic portrait, not a customer photo.

The photo used as input for Childhood Me — A real run of this skill. The input is a synthetic portrait, not a customer photo.

Ready to use

Childhood Me

See yourself as a kid — 4 ages from one photo

Use this skill

Your photo: Required
Output: 4 images, square 1:1
Category: packs
Cost: 4 credits

childhood
de-age
nostalgia

What it does

What you get back

Creates four individual square doodle images from a single adult photo. Each image shows the same recognizable person at a different childhood age (3, 6, 10, and teen) while preserving their core identity features — hair colour, eye colour, face shape, and distinguishing marks. Age-appropriate proportions, clothing, and features change per stage, but the character is unmistakably the same person across the set.

When the agent picks it

Use when the user wants to see themselves as a child, wants a de-aged doodle, childhood version, or asks for "me as a kid" from an adult photo.

The full skill file

Every skill is one SKILL.md — this is the exact file the agent reads.

src/mastra/skills/childhood-me/SKILL.md

Childhood Me

Turn one uploaded adult photo into 4 separate square doodle portraits — the same person de-aged to four childhood stages, each its own image file.

When this is the right skill

Pick this skill when the user wants:

A childhood version of themselves or someone else
“Me as a kid” / “what did I look like as a child”
De-aged portraits at multiple stages
A nostalgia / throwback set from an adult photo
“Show me growing up” or “baby me to teen me”

Do NOT use this for a single baby photo (use normal doodle), for aging someone UP (not supported), or for turning a child photo into an adult.

What to draw

Four separate 1:1 images, each a head-and-shoulders naive marker-and-ink doodle of the same person de-aged to one childhood stage:

Age 3 — toddler: very round chubby cheeks, huge head-to-body ratio, button nose, wide-set oversized eyes, baby-plump jawline. Simple toddler outfit (onesie or elastic shorts).
Age 6 — early school: round cheeks with a gap-tooth grin (missing front teeth), head still large, slightly longer face. Simple graphic tee and light-up sneakers.
Age 10 — pre-teen: face lengthening, adult front teeth slightly oversized for the face, less baby-fat, emerging jawline. Hoodie and chunky trainers.
Teen — around 14: proportions nearing adult but softer and less defined, rounder jaw, narrower shoulders. Oversized band tee and canvas sneakers.

Identity consistency is paramount: hair colour and pattern, eye colour, skin tone, face shape, freckles/moles, and glasses persist across all four. Only age-appropriate proportions, clothing, and developmental features change. No adult attributes (facial hair, makeup, jewellery) on any child.

How to run it

Confirm a photo is attached. If none is available, ask the user to attach one.
Call generateDoodle with skill: "childhood" and the uploaded photo as imageUrl.
Pass refImageUrl when the message includes a separate style reference.
The tool returns FOUR separate images (one per age). Cost: 4 credits (one generation per image).
Report the result briefly; the app renders all four images in a row.

Following up

Offer to regenerate a specific age the user dislikes, suggest adding a description for era-specific clothing (”80s kid”, ”90s kid”), or recommend the seasonal-pack skill for a different multi-image concept.