Add Speech Bubble

Catalog

• Display name: Add Speech Bubble
• Categories: Social Media, Layout & Design
• Status: Ready
• Route label: GPT Image 2 overlay
• Tone: blue
• Short description: Add exact dialogue bubbles at the right subjects.

What this skill does

• Adds oval, cloud, rectangle, starburst, caption, or text-only bubbles.
• Binds each text string, body, and tail endpoint to a declared speaker.

How to use

Provide image, exact text per speaker, style, and target or location.

Add dialogue

Add an oval bubble above the woman, tail at her mouth, exact text: “五分钟后见！”

Workflow

1. Map each exact line to speaker, tail target, placement, style, and reading order.
2. Reserve faces and important objects; choose fill, stroke, font, and contrast.
3. Use gpt-image v1.0, GPT Image 2 Medium, original ratio, batchCount: 1.
4. Preserve source outside bubble regions; do not invent dialogue or cover faces.
5. Verify copy character by character, tail endpoint, speaker, layer order, legibility, and bubble count. When this Skill selects GPT Image 2, use quality: "medium" and imageSize: "2K" by default; change either only when the user explicitly requests another tier.

Every GPT Image 2 call in this Skill defaults to Medium at 2K; use another tier only when the user explicitly requests it.

User-facing output

• Media type: One annotated image
• Default quantity: 1
• Content per image: Source plus requested bubbles only
• Default layout: Original frame with overlays
• Model policy: GPT Image 2 Medium with agent-led copy and placement verification
• Downstream use: Comics, social posts, and annotations