About Prompt Workbox

Prompt Workbox helps LLM developers, RAG engineers, and prompt authors craft, test, and organize prompts without sending text to a cloud prompt platform.

Core tools run entirely in your browser. Drafts, custom templates, and version snapshots stay in localStorage on your device. Optional blog content is published separately at /blog.

Your prompts never leave your browser.

Principles

• Local-first: no account, no cloud sync, no default AI rewrite
• Focused on edit, template, clean, compare, lint, and export
• Token counting as a companion — not the whole product
• Works with MDWorkbox and TokenWorkbox in the same toolkit series

Core tools

• Prompt Editor — text & ChatML with live token counts
• Templates — built-in and local custom templates with variable placeholders
• Cleaner — normalize, strip markup, escape, or compress prompts

Workbox matrix

Use MDWorkbox to clean Markdown source material, assemble prompts here, then send finished prompts to TokenWorkbox for Exact token counts, cost estimates, and RAG chunking.