English, Choose language
Menu
English, Choose language

If your AI agent, the program that does work for you on the computer, also answers a simple question with a wall of text, you're not alone.

Where you are now

What if length isn't the real problem? “Write less” doesn't fix it, because an agent that just shortens may also cut what it must not cut.

Meanwhile every reply like that costs you time. You read a page to find one sentence, dozens of times a day, and the failed check sits in paragraph four, after three paragraphs of things that worked.

And after a week of long answers you stop reading to the end, and that is exactly when you miss the line that mattered.

Axel Jedlitschka

Who's talking: Axel Jedlitschka

It all started with an income and expenses table I wanted to build for myself in Excel. After a few versions I moved to Google Sheets, and from there it kept growing.

In my phone sales jobs those skills really served me, because I built dashboards, screens that gather the numbers in one place, good-looking and effective. From there I started building systems and automations, processes that run on their own, for business owners. A good dashboard works the way an agent's reply should: you look at the signal and know whether you need to act.

Axel Jedlitschka

What to expect

  • One token on top

    Know from the first line whether it's done, whether the agent needs something from you, or whether it's still working.

  • Bad news up front

    Get failures and risks every time, and first, even in a one-line reply.

  • A location, not a recap

    Get the location of the file with the result, instead of a summary of what the agent did.

  • One question

    Get one question with three options, not an eight-step guide for something you already have.

  • “Done” only with proof

    Know that when it says done, there's a passing check, the output of a real run, or a before/after comparison under it.

There's a way

“Write less” isn't enough. The fix is a fixed shape, with a clear rule for when to go beyond it, so you always know where to look. The shape has four parts.

  1. A status line.The first line of every reply is one token and one sentence: ✅ done and verified, ⛔ blocked and the agent needs something from you, ▶️ still working or waiting for your approval.
  2. A 6-line budget.A reply is at most 6 lines, status line included, and code, a table or a document you asked for doesn't count. The agent goes longer only in one of five cases, and names it: gate failure, deviation, ratification, risk or explicit request.
  3. What stays out.No “Great question”, no summary of what the agent did, no story of the process and no repeating your request back to you. Recommendations you didn't ask for? At most one final line starting with “Suggestion:”.
  4. When something is missing.First the agent tries to get it itself: a file, a tool, a search. If it can't, it asks one question with three options: “I have it”, “I don't, give me steps”, “Skip”. Only after “I don't” does a table of at most 5 steps arrive.

The budget never hides bad news.

The way, in one kit

Reply kit · a kit by Axel Jedlitschka

Every reply from the agent: a status line, the result, and nothing you didn't ask for.

  • The install script shows what it will do, and asks before writing anything
  • The block sits between two marker lines, so it is never added twice
  • A copy of your file is kept as it was before the change
  • A skill for Claude Code: an instruction file the agent loads when it's relevant
  • The contract block for CLAUDE.md, the instruction file the agent reads, so the contract applies to every reply
  • The same block, ready to paste into any other agent, like Codex, Cursor or Gemini CLI
  • Templates in English, Hebrew and Spanish
  • Six before/after examples, including a task blocked on a password and a risky action
  • A checklist of 21 checks, to review a reply before it goes out
  • Install in one command, which asks before writing and keeps a copy of your file
The first page of the Reply kit guide

The comparison

Free

Reply kit

What's on the first line
One token and one sentence
Where the bad news is
Always reported, and first
When something is missing
One question with three options

A reply with no rules

What's on the first line
“Great question!”
Where the bad news is
In paragraph four
When something is missing
An eight-step guide

The offer

A skill, a block for any agent, templates in three languages, a checklist, and a 7-page guide

Price: FreeNo credit card, and no part is held back for payment

Get the kit

The kit is free and you get it right away. You only get marketing messages if you tick the box, and you can unsubscribe at any time.

Right after you leave your details, a thank-you page opens with a button to download the kit. You choose the language of the guide in the form: Hebrew, English or Spanish.

By Axel Jedlitschka

What you get

Promised

  • The full kit, free, as soon as you leave your details.
  • No part is held back for payment.
  • Your details are not sold, and are not passed to anyone except the infrastructure providers that run the site.
  • A question about the kit? Send a message on Instagram to @theonlyaxelj.

Not promised

  • A result. The kit gives the agent a fixed shape for every reply, and the result depends on how you apply it and on your business. It's general knowledge, not personal advice.
  • That the contract applies without the block. The skill loads only when relevant, and the block in your instruction file is what makes the contract apply to every reply.

There is no purchase and no payment here, so there is nothing to cancel. You can withdraw your consent to marketing messages at any time.

Questions?

How do I get the kit?

You leave your details, and a thank-you page opens right away with a button to download the kit.

Is it really free?

Yes. No credit card, and no part is held back for payment.

I'm not a programmer. Can I install it?

You unzip the file you downloaded, and from the folder that opens you run one command in the terminal, the window where you type commands to the computer: bash install.sh. The script shows what it will do and asks before writing anything. The default answer is “no”.

Isn't 6 lines too few?

For most replies it's plenty, because code and tables you asked for don't count. When more is needed, there are five defined cases, and there's the escape word: write “expand”, and the next reply runs as long as it needs.

Won't the agent hide problems to stay short?

That is exactly what the contract forbids. Failures and risks are always reported, and first, even in a one-line reply.

I already have an instruction file for my agent. Will this overwrite it?

No. The block is added at the end, and a copy of the file as it was is kept next to it. If you have other rules about reply length in there, remove them, or the agent gets two contracts that contradict each other.

Will I get spam or phone calls?

Your details are used to send you the kit and to follow up on it. You only get marketing messages if you ticked the consent box, and you can unsubscribe at any time. You can also ask to see, correct or delete your details.

Where do I get help?

A question about the kit: send a message on Instagram to @theonlyaxelj.

What now

You can keep reading a wall of text to find one sentence, or you can know from the first line whether the ball is in your court.

A kit to install and a 7-page PDF guide: a status line, a 6-line budget and a block for any agent, free.

Price: FreeNo credit card

Get the kit

The kit is free and you get it right away. You only get marketing messages if you tick the box, and you can unsubscribe at any time.

The kit is ready to download as soon as you leave your details

The kit is free and you get it right away. You only get marketing messages if you tick the box, and you can unsubscribe at any time.

Your details are used to get back to you. What is collected, why, and who receives it: see the Privacy Policy (opens in a new tab).