Couple Doodle 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 Couple Doodle — A real run of this skill. The input is a synthetic portrait, not a customer photo.

Ready to use

Couple Doodle

Turn a photo of two people into one illustrated couple portrait

Use this skill

Your photo: Required
Output: Single square 1:1
Category: avatars
Cost: 1 credit

couple
portrait
two-people
relationship

What it does

What you get back

Takes a photo of two people and redraws them as a single illustrated couple portrait in naive marker-and-ink doodle style. Preserves each person's individual likeness — hair, face shape, height difference, skin tone, glasses — and keeps their physical relationship from the photo (holding hands, arm around shoulder, leaning together). Supports optional occasions like anniversary, engagement, proposal, wedding, or Valentine's that add matched doodle embellishments.

When the agent picks it

Use when the user wants a doodle of two people together — "us as cartoons", "me and my girlfriend", "draw me and my partner", "couple portrait", "me and my boyfriend as a cartoon", "my partner and me", "anniversary", "engagement doodle", "our wedding cartoon". Requires a photo with BOTH people visible. For a single-person avatar use doodle-avatar instead; for a greeting card (birthday, thank-you) use gift instead.

The full skill file

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

src/mastra/skills/couple-doodle/SKILL.md

Couple Doodle

Turn one uploaded photo of two people into an illustrated couple portrait.

When this is the right skill

Pick this skill when the user wants a doodle of two people together in a single portrait — partners, spouses, best friends photographed side by side. The photo must contain both people.

Choose a different skill instead when the user asks for:

a single person’s avatar → doodle-avatar
a greeting card with a message → gift
multiple panels or expressions → doodle-collage
head-to-toe or action poses → full-body-collage
a sticker sheet → sticker-pack
mood-word captioned panels → mood-captions
artwork with no photo at all → surprise-me

What to draw

Redraw both subjects as illustrated doodle characters in the house style: bold clean outlines, simplified features, flat cheerful colors, playful proportions. Likeness is the hard requirement — see references/two-subject-likeness.md for the full anchor checklist.

Preserve from the source photo:

each person’s own hair colour, style, and length
each person’s face shape, skin tone, and distinguishing marks
glasses, facial hair, piercings, and accessories per person
relative height and body-size difference between the two
their left/right positions exactly as photographed
their physical relationship (holding hands, arm around shoulder, foreheads together, etc.) — never invent a generic pose

When an occasion is detected from the user’s message (anniversary, engagement, proposal, wedding, Valentine’s, just-us), add small occasion-matched doodle embellishments around the couple — the same pattern the gift skill uses for its occasions.

How to run it

Confirm a photo is attached and that two people are clearly visible. If only one person is in the photo, tell the user: “I can see one person — for a couple doodle I need a photo with both of you in it. Want to upload a different photo, or shall I make a single avatar instead?”
Call generateDoodle with skill: "couple", the uploaded photo as imageUrl, and the user’s occasion or message as description (if any).
Pass refImageUrl when the message includes a separate style reference.
Report the result briefly; the app renders the image itself.

Following up

After a successful generation, offer one next step — a greeting-card version (gift skill), a collage of the couple in different moments, or a refinement such as a different theme or occasion.