slopfree.ca

I'm over the text wall.

“Great question! It’s not a human, it’s AI! Let’s delve in. Blah. Blah. Blah. A literary flourish here, a tidy list of three there, and a recap of everything you just read. No point anywhere. I hope this helps!”

Let’s remember who we’re writing for.

golden-prompt.md
Your reader is smart and busy. Give them the whole answer in the fewest words.
- Lead with the answer.
- Keep every fact; drop every word that carries none.
- Plain words, short sentences, no filler or flourishes.
- Unsure? Say so once. Never make things up.
Golden prompt 1 · paste it at the end of any prompt

About

Same meaning. At least 20% less text.

Slop Free is lossless compression for AI writing. A prompt only becomes golden if machines can show it keeps the facts, and people can show they understand it at least as well.

  1. 1

    Eval

    Every week, candidate prompts summarize 10 random Wikipedia articles. One passes only if it keeps 98% of the points a plain "Summarize this" keeps, adds no errors, and uses at least 20% fewer words.

  2. 2

    Read

    The winner faces the golden prompt in the quiz. You read short summaries from both and answer one question each. You won't know which wrote what.

  3. 3

    Rank

    If readers understand the challenger at least as well and read it faster, it becomes the new golden prompt. If not, it goes in the losers folder with why.

Why Wikipedia

Random articles on every subject, with a pinned revision, so each prompt is judged on exactly the same text.

Close but no cigar

Every losing prompt is kept with the reason it lost. Browse the losers or submit your own.

No tracking

We store your answers and reading times. No accounts, no cookies, no IP addresses.

Models we use

JobModel
Writes the summariesgpt-6-luna
Judges which points were keptgpt-6.2-sol
Drafts new candidate promptsgpt-6.2-sol

A prompt that works on one model can fail on another, so every result on this site names the models that produced it.

How to use it

Paste it at the end, or save it once.

The golden prompt is a postfix: text that goes after your request. Models pay close attention to the end of a prompt, so that's where it does the most work.

For one message

Write your request as usual, leave a blank line, then paste the prompt.

Explain how a heat pump keeps a house warm at -20°C.

Your reader is smart and busy. Give them the whole answer in the fewest words.
- Lead with the answer. ...

For every chat

Save it in your assistant's custom instructions and every reply follows it, without pasting.

Where to save it

ChatGPT
Settings → Personalization → Custom instructions
Claude
Settings → Profile → personal preferences, or a Project's instructions
Gemini
Saved info, or a Gem's instructions
API and agents
The end of your system prompt

Menu names change between app versions. If you can't find these, look for "custom instructions" or "personalization".

Tips

  • Keep it last. Put task-specific instructions before it, so the golden prompt is the final thing the model reads.
  • Re-paste in long chats. Models drift back to old habits as a conversation grows. Pasting it again resets them.
  • Edit freely. Drop a rule you disagree with. Just know the quiz ranks the version shown here, not your edit.

Quiz

Help train the golden prompt.

Answer just 5 short questions to contribute. Read each summary at your normal pace.

Human First Design

Crafting for humans, with AI

  1. Design for the human first.

    Most end users are people, and forcing them to sort through a wall of AI output is a bad experience.

  2. Protect the negative space.

    Leave things out so the real point stands out. Human attention is small, and AI output is big. Edit. Edit. Edit.

  3. The burden is on the designer.

    You, as the designer or writer, hold the context. It’s your responsibility to make the output understandable. Do not push the responsibility onto the reader to decode a confusing design or sift through a wall of text to find the point.

  4. Make it obvious.

    Each paragraph should make its point obvious, as should each next step in an application. A user or reader shouldn’t have to guess what’s going on, and it should be easy to tell when something is wrong. Errors should give enough information to correct the problem, or at minimum enough to support troubleshooting.

  5. Peel the onion.

    Show the essence first, then give the user a path to dig deeper. In writing, lead with the most important points and follow with the necessary details.

  6. Respect their time.

    Just because we can create at light speed doesn’t mean anyone has to consume it. It’s still the user’s or reader’s choice. You have to earn the right for someone to spend time reading your doc or using your system.

  7. Never accept the first output.

    Treat the first output, whether text or an application, as a draft. Edit it heavily using the principles above. Remember: put the human first.