Skip to content
TakoAPI
Back to skills

nano-banana-cog

by nitishgargiitd

Coding Agents & IDEs
—Not yet rated
Your rating

Brief

Nano Banana × CellCog.

Description

What This Skill Does

Wraps Nano Banana image generation through CellCog's execution layer. Supports text-to-image, character-consistent series, multi-image composition, and photo editing at 1K–4K resolution. Any OpenClaw agent can invoke it with a single prompt.

Exposes Nano Banana's character consistency and multi-resolution output inside any OpenClaw agent without a separate image API integration.

When to Use It

  • Generating product shots for e-commerce listings
  • Creating a brand mascot in multiple poses and contexts
  • Building a visual story sequence with a consistent character
  • Transforming photos with style transfer or background swaps
  • Producing photorealistic marketing scenes from text descriptions

Example Workflow

INPUT

User asks: Create a tech entrepreneur character, then show them presenting on stage and working at a laptop

AGENT
  1. 1

    Calls CellCog client with chat_mode set to 'agent'

  2. 2

    Sends detailed character description including age, style, and lighting for the initial portrait

  3. 3

    Receives first image with character identity established

  4. 4

    Sends follow-up prompts referencing 'the same character' for the stage and laptop scenes

  5. 5

    Returns all three images with consistent character appearance

OUTPUT

Three photorealistic images of the same character across different professional contexts

Installation

clawhub install nitishgargiitd/nano-banana-cog

Run the CLI command, or paste the prompt into Claude Code, Cursor, Codex or whichever coding agent you use.

Links