diagram

Generate publication-quality technical diagrams from natural language prompts.

3|Updated Jan 7, 2026
One-click install
npx skills add https://github.com/flight505/nano-banana --skill diagram-flight505
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: diagram
Source: https://github.com/flight505/nano-banana/tree/main/skills/diagram
Command: npx skills add https://github.com/flight505/nano-banana --skill diagram-flight505

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires python-dotenv, and includes scripts (resource) components.

What problem does it solve?

This Skill automates the creation of publication-quality technical diagrams from natural language prompts, reducing manual drafting time and ensuring professional visuals.

Core Features & Use Cases

  • Smart Iteration: Regenerates only when the diagram quality is below threshold, saving time and API usage.
  • Document-Type Aware: Applies different quality standards for architecture, journal, presentation, and other document types.
  • AI Quality Review: Gemini 3 Pro evaluates technical diagrams for accuracy, clarity, and layout.
  • Use Cases: Architecture diagrams, flowcharts, data-model visualizations, and infrastructure diagrams for software engineering, architecture decisions, and technical documentation.

Quick Start

Use the diagram skill to generate a diagram from a natural language prompt, for example:

  • python skills/diagram/scripts/generate_diagram.py "Microservices architecture with API gateway" -o architecture.png --doc-type architecture
  • python skills/diagram/scripts/generate_diagram.py "User authentication flow" -o flow.png --doc-type presentation

Frequently Asked Questions about diagram

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

FAQPage Schema
How do I generate an architecture diagram from a natural language prompt?

To generate an architecture diagram from a natural language prompt, you run the Python script and specify your text prompt alongside the desired output file name. The Skill parses your prompt to produce a publication-quality technical diagram automatically.

Can I use this to create a presentation-ready flowchart?

Yes, you can create a presentation-ready flowchart by specifying the document type as presentation. The generation process applies different visual quality standards based on the target document type to ensure professional results.

How does the AI quality review process work for technical diagrams?

The AI quality review uses Gemini 3 Pro to evaluate generated technical diagrams for accuracy, clarity, and layout. It applies a smart iteration mechanism that only regenerates the diagram if the initial quality falls below a defined threshold.

Do I need an OpenRouter API key to generate technical diagrams?

Yes, you need OpenRouter API access configured in your Python environment using python-dotenv. The Skill relies on this external API access to parse your natural language prompts and produce the final diagrams.

What types of technical diagrams can I create from text prompts?

You can create architecture diagrams, flowcharts, data-model visualizations, and infrastructure diagrams. These visualizations are suited for software engineering, architecture decisions, and technical documentation contexts.

Are there limitations when generating publication-quality diagrams via AI?

The main limitation is the dependency on external API usage and the quality threshold set by the AI review. Complex prompts may require multiple iterations, consuming more API calls before reaching the desired publication-quality standard.