@rewdy/chatty-caddy
v0.2.0
Published
A CLI notebook for storing and running AI prompts
Maintainers
Readme
chatty-caddy
A CLI notebook for storing, browsing, and running your frequently used AI prompts.
Prompts are saved as Markdown files in ~/.chatty-caddy with YAML frontmatter — human-readable, easy to back up, and version-controllable.
Installation
npm install -g @rewdy/chatty-caddyUsage
Both chatty-caddy and the shorter alias chaca work for all commands.
Add a prompt
chaca addWalks you through a prompt to enter a label, description, and body. The prompt is saved immediately to ~/.chatty-caddy.
Browse & manage prompts
chaca listOpens an interactive browser with fuzzy search. From there you can:
- View the full prompt body
- Run the prompt directly in Claude, Codex, or GitHub Copilot
- Edit any field
- Delete a prompt
Help
chaca helpPrompt storage
Each prompt is saved as a Markdown file at ~/.chatty-caddy/<uuid>.md:
---
id: 3f2a1b4c-...
label: Refactor to functional
description: Converts a class component to a functional React component
createdAt: 2026-04-21T10:00:00.000Z
---
Refactor the following React class component into a functional component using hooks...Adding AI tools
Supported tools (Claude, Codex, GitHub Copilot) are defined in src/lib/tools.ts. Adding a new one is a single array entry:
{
id: "my-tool",
label: "My Tool",
command: "my-tool-cli",
inputMethod: "arg", // or "stdin"
}Development
Requires Bun.
bun install
bun run dev add
bun run dev listTo build a standalone binary:
bun run build
# outputs dist/chatty-caddy