tt-a1i/archify turns a plain-English system description into a self-contained HTML diagram with multi-format image export. It provides a CLI for validation and rendering, and a GPU is not required.
Project overview
It converts text descriptions directly into polished architecture, workflow, sequence, data-flow, and lifecycle diagrams while mapping component names to correct visual categories without requiring an icon library.
Project type
Infrastructure · Prompt Engineering
Use cases
Knowledge Q&A
Deployment
Refer to project documentation
License
MIT
Best for
Developers and AI engineers needing to generate self-contained HTML or image diagrams from plain English descriptions via a CLI.
Key capabilities
Creates architecture, workflow, sequence, data-flow, and lifecycle diagrams from plain English descriptions.
Copies PNG to clipboard or exports crisp up-to-4x images in PNG, JPEG, WebP, or SVG formats.
Maps component names like aws.lambda or postgres to the correct visual category without requiring an icon library.
Exposes a standalone CLI for running doctor checks, demos, rendering, validation, and artifact checking.
Limitations and risks
Oversized diagrams automatically step down in resolution or size to stay within browser canvas limits.
The tool is not a WYSIWYG application or a general-purpose drawing editor.
Getting started
Installation is rated easy and is achievable through a single npx command. Run `npx skills add tt-a1i/archify -g` to install.
Evidence and sources
GitHub project description: Any agent Skill: generate beautiful architecture diagrams with dark/light theme toggle and PNG/JPEG/WebP/SVG export
README: Archify is an agent skill for Claude, Codex CLI, and opencode. It turns a plain-English description of your system or process into a polished, self-contained technical diagram
README: npx skills add tt-a1i/archify -g
README: SVG | READMEs, blogs, Figma, Illustrator, and lossless scaling