Will McGugan's Toad Terminal AI

Table of content

Will McGugan spent twelve years as a freelance Python contractor before creating Rich and Textual, two Python libraries downloaded over 3 billion times. When major tech companies started shipping terminal-based AI coding agents with flickering displays and broken interactions, he built Toad to fix what they got wrong.

Background

Created Rich (55k GitHub stars), a terminal formatting library used by pip
Built Textual (34k stars), a Python TUI framework
Founded Textualize in 2021 with seed funding in Edinburgh
Python Software Foundation Fellow and author of “Beginning Game Development with Python”
Currently on sabbatical after closing Textualize in mid-2025
@willmcgugan on Twitter

McGugan built the Toad prototype “across two afternoons in a nerdy caffeinated rage while listening to metal music” after trying Claude Code and Gemini CLI.

What Toad Fixes

McGugan watched AI coding tools from Google and Anthropic ship with problems his libraries solved years ago:

Problem	Existing Tools	Toad’s Solution
Screen flickering	Redraw entire screen on updates	Update partial regions (even single characters)
Text selection	Copies formatting characters and line numbers	Clean text selection and copying
Terminal resize	Output becomes garbled	Graceful resize handling
Interactive apps	Many tools break on TUIs	Full color, interactivity, mouse support
Scrollback	Content becomes read-only	Accessible and interactive after scroll

Installation

curl -fsSL batrachian.ai/install | sh

Or with UV:

uv tool install -U batrachian-toad --python 3.14

Supported AI Agents

Toad uses the Agent Client Protocol (ACP) to provide a single UI for 12+ coding agents:

OpenHands
Claude Code
Gemini CLI
Aider
And more through ACP

Start a coding session:

toad

Run as a web service:

toad serve

Key Features

Fuzzy file search. Type @ to search files with .gitignore filtering.

Streaming Markdown. Renders large documents without performance loss, with syntax-highlighted code blocks.

Jupyter-style navigation. Move through conversation blocks for easy content interaction and export.

Command palette. Access all shortcuts via Ctrl+P.

Design Philosophy

From his announcement post:

“I wanted the conversation with AI to feel like a natural extension of a traditional terminal-based workflow. Most tools stop at displaying monochrome output from commands.”

McGugan emphasizes terminals as underutilized platforms. While most developers treat terminals as dumb output streams, his libraries prove they can host responsive, interactive applications.

The Textualize Experience

McGugan ran Textualize for three years before concluding that “Textual has always been a solution in search of a problem.” The company built impressive software but couldn’t find a sustainable business model.

What he kept:

Rich and Textual remain maintained open source
The terminal UI expertise directly informed Toad
Community engagement via Discord continues

Building in Public

McGugan’s approach to Rich:

Released early, shared progress publicly on Twitter
Implemented features requested by multiple users
Willing to bump major versions for API elegance
Documented breaking changes transparently

Progress bars weren’t originally planned but became one of Rich’s most popular features after repeated community requests.

Key Takeaways

Principle	Implementation
Fix real problems	Built Toad after using broken AI tools
Use existing expertise	Terminal UI knowledge from Rich/Textual
Protocol over implementation	ACP enables multi-agent support
Release early	Share progress, iterate with community
Open source first	AGPL-3.0 license, seeking sponsorship

Links

Next: Jesse Vincent’s Superpowers Framework