Skip to content
  1. Skills at a glance
  2. Understand you first
  3. /pstack:sharpen

/pstack:sharpen

Type it yourself

In one lineTightens your prompt: adds the steps and checks an expert would think of. It doesn't do the work.

When to use
An important job, and you want the request right before you send it.
Say it like this
/pstack:sharpen Write a migration guide for the v2 config format.

Just asking vs. using it

The same request, typed plainly or with /pstack:sharpen. Here's what changes.

Just askingWith /pstack:sharpen
Makes the task bigger by piling on more work.First says what a top professional would do that you didn't ask for.
Starts from material someone already digested for you.Names the rawest material a professional would start from.
Counts the result right because you and the AI like it.Names something outside you and the AI that would show it's right.

What the document says

About 98 words, 1 min to read.

OpeningBefore rewriting your prompt, the AI names what a top professional would add, the rawest material, the methods and tools, and an outside check. Then it rewrites the prompt to ask for exactly that and leaves your choices open.

See the wording worth borrowingRead the full text

Wording worth borrowing

A few sentences from the original: what each one makes the AI do, why it works, and how you can borrow it.

  1. Name the direction you don't want

    I want a deeper version of this task, not a bigger one.

    Makes the AI
    It rewrites your prompt to go deeper, instead of piling more work into it.
    Why it works
    "Deeper" and "bigger" both mean "more"; putting them side by side tells the AI which way to add.
    Borrow it
    When you ask for an improvement, name the direction: "I want this article sharper, not longer."
  2. Borrow an expert's eyes

    Say what a top professional in this field would do here that I'm not asking for.

    Makes the AI
    It adds the steps you didn't think of but an expert would never skip.
    Why it works
    "Do better" has no direction; "what would a top professional do" gives the AI someone concrete to copy.
    Borrow it
    Before an edit, ask: "What would a senior editor do to this draft that I haven't asked for?"
  3. Ask for an outside judge

    Say what, other than you or me, would show the result is right.

    Makes the AI
    The rewritten prompt includes an outside check, such as real data, a test, or actually following it.
    Why it works
    The writer and the asker both tend to like the result; naming something outside them is what makes it convincing.
    Borrow it
    Add: "Besides you and me liking it, what would prove this migration guide is right?"
  4. Mark which calls are yours

    Leave choices that are mine to make open instead of deciding them.

    Makes the AI
    It leaves your decisions, like budget or style, to you instead of making them.
    Why it works
    Saying which decisions are yours stops the AI from making them just to look thorough.
    Borrow it
    Write: "I'll pick the colors; list two or three options and don't choose one."

Full text

The SKILL.md the AI reads, as written.

descriptionMake a task deeper, not bigger, then rewrite the prompt. Use for /pstack:sharpen.

I want a deeper version of this task, not a bigger one. Before rewriting it, do four things:

  1. Say what a top professional in this field would do here that I'm not asking for.
  2. Name the rawest material they would start from instead of what has already been digested for me.
  3. Name the specific methods and tools they would use.
  4. Say what, other than you or me, would show the result is right.

Then rewrite my prompt to ask for exactly that.

Leave choices that are mine to make open instead of deciding them.

Original by Lauren Tan (poteto), MIT license. Checked against the pstack mirror at 97de40d. View on GitHub

An unofficial beginner's guide to pstack, not its official docs. pstack is made by poteto; the official repo is the source of truth for commands and behavior. The original guide source is on GitHub.

Hosted by airali. Adapted from Xiaoxiaodong's MIT-licensed guide.