effort-free
Free, for anyoneNo sign-upNo data leaves your screen

Think better.
Keep your
data yours.

Free tools that turn rough ideas into prompts experts would write. Copy. Paste. Done.

The tools see nothing.
The tools run in your browser. Nothing leaves your browser.
Works with Copilot · ChatGPT · Claude · Gemini
The Zero-Data Promise

Your words stay yours.

Your text never touches a server. The browser hides names and numbers before you send. No database stores your words.

Step 1
Pick a tool here
You open the page. You choose the task. A playbook or prompt appears for the task.
Runs in your browser
Step 2 · Your turn
You paste it into your own tool
Copilot at work. ChatGPT at home. Claude. Gemini. Whatever you trust. Your company's data stays inside the tool your company already approves.
This step is private
Step 3
You keep the answer
The output stays with you. Nothing to show anyone. Nothing to leak. Nothing to sell. That is the whole point.
Your work, your file
01 · No upload
Your files stay put. The site never asks for them.
02 · No training
Your words do not train any model here.
03 · No storage
The site holds nothing. No log-in means no profile.
04 · No catch
Always free. No upsell on the next page.
See it work · Live demo

A rough thought in. A clean prompt out.

Pick the person who sounds like you. Watch The Architect turn a messy idea into a prompt you can paste anywhere.
The Architect - Live Preview
Zero-Data
Executive · raw thoughtStep 01
Structuring your zero-data prompt…
Architect · ready to pasteStep 02
Built in your browser. The raw thought above never leaves this page.
When to use what · three pillars

Eight tools. Three clear jobs.

Tools are grouped by what you're trying to do. Start with the outcome you want.

Every tool below runs the same four steps. Frame. Validate. Execute. Refine. Each page labels the step so you know where you are.

Pillar 03
Frameworks & Library.
Borrow mental models from history and peers.
When the answer isn't a draft, it's a better way to think. Borrow the best ideas from history, nature, and the working professionals already using them.
New here?

New to AI? Start here.

A zero-jargon guide to taking your first step. No setup, no accounts, no assumptions about what you already know.

Start the guide

The engine behind the tools.

INPUTtextCALL 1picksframeworkcraftdomainLIBRARYCALL 2writesOUTPUTprompt

Two passes, not one

Most tools send your input to one AI call. This site uses two. The first call picks which knowledge fits your question. The second call writes the prompt using that knowledge. Haiku runs both. The library does the thinking.

Two passes, not one

Splitting the work keeps each call focused. The first call picks. The second call writes. Haiku is built for narrow tasks. Asking one call to choose knowledge and write prose makes both worse. Two passes cost a bit more. The output is sharper.

FRAMEWORKSCRAFTDOMAINFAILURE70 chunks across 4 tiers

A curated library

A library of frameworks, mental models, and hard-won lessons sits behind every tool. Every prompt draws from it. You ask a question. The system pulls patterns that fit. You get back a prompt experts would have written.

A curated library

The library holds 70 chunks across four tiers. Frameworks. Craft. Domain. Failure modes. Each chunk is 200 words of distilled practice. No vector search. No magic. A small catalog picked with care beats a giant one searched blind.

BEFOREAlexAcmealex@acme.io07700900123[NAME_1][CO_1][EMAIL_1][PHONE_1]tokens:0time:0.0sREDACTED[NAME_1][CO_1][EMAIL_1][PHONE_1]NO LOG

Nothing leaves your computer

The browser redacts your text before it leaves your computer. Names, emails, and numbers turn into placeholders. The API sees the placeholders. The tracer records timing and token counts, not content. No user database exists. The site keeps no record.

Nothing leaves your computer

Redaction runs client-side in pure JavaScript. Fifteen detectors catch emails, phone numbers, names, card numbers, and more. You see every placeholder before you send. Anthropic holds traffic for 30 days to catch abuse, then deletes it. The site itself stores nothing.

token costCall 12345678~10% eachCACHED CALLSFIRST

Why this stays free

The system prompt that teaches Haiku how to behave is long. Without caching, every call would pay for those tokens. Prompt caching pays for them once. Every later call costs about 10% of the first. That is why the site stays free.

Why this stays free

The system prompt is about 3,000 tokens. It defines identity, contract, principles, and injection defense. Each tool has its own. Anthropic caches the prompt for five minutes. During a busy hour, the cache stays warm. That drops per-call cost from pennies to a fraction of a penny.

About

Jose Ortiz Flores built this site. The site shares the prompt patterns that make AI useful for real work and explains all AI related things to help everyone learn.

The Zero-Data Promise
Your data never leaves your screen.
01 · No upload
Files stay put.
02 · No training
Your words, your own.
03 · No storage
No logs. No profile.
04 · No catch
Always free.