Nano Banana–like Local Image Reasoning Skill

Parses prompts into structured multi-stage image generation plans for ComfyUI execution.

7|2|Updated Oct 28, 2025
One-click install
npx skills add https://github.com/SkastVnT/AI-Assistant --skill nano-banana-like-local-image-reasoning-skill
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: Nano Banana–like Local Image Reasoning Skill
Source: https://github.com/SkastVnT/AI-Assistant/tree/main
Command: npx skills add https://github.com/SkastVnT/AI-Assistant --skill nano-banana-like-local-image-reasoning-skill

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It enforces a planner-driven approach for local image generation, ensuring explicit planning before any generation to maintain consistency across panels, characters, and props.

Core Features & Use Cases

  • Plans multi-panel stories from user prompts and preserves character continuity across frames.
  • Orchestrates ComfyUI-based execution with detector/inpaint loops to improve output fidelity.
  • Use case: transform a prompt into a structured panel sequence with defined portraits, scenes, and overlays.

Quick Start

Plan a multi-panel comic from a user prompt and execute it with ComfyUI to ensure character and prop continuity.

Frequently Asked Questions about Nano Banana–like Local Image Reasoning Skill

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I maintain character continuity across multiple image generation panels in ComfyUI?

To maintain character continuity across multiple image generation panels, this Skill enforces a planner-driven approach that parses prompts into a structured, multi-stage plan before executing ComfyUI detector and inpaint loops for panel-level consistency.

What is a planner-driven local image workflow for multi-panel comics?

A planner-driven local image workflow for multi-panel comics creates an explicit structured plan defining portraits, scenes, and overlays before any generation begins, ensuring character and prop consistency throughout the ComfyUI execution sequence.

Do I need local image pipeline components and models installed to use this ComfyUI planning Skill?

Yes, you need local image pipeline components and related models present in your environment, because this Skill orchestrates ComfyUI-based execution relying on detector and inpaint loops for accurate local image generation planning.

How do I plan a multi-panel story sequence from a single text prompt?

To plan a multi-panel story sequence from a single text prompt, the Skill parses your input into a structured panel sequence with defined portraits, scenes, and overlays to preserve character continuity across frames.

When should I use detector and inpaint loops for local image generation?

You should use detector and inpaint loops for local image generation when you need to improve output fidelity and guarantee panel-level continuity in multi-panel stories orchestrated through the ComfyUI execution engine.