Skip to content
Back to skills

One Line

ASecurity

Explain it in exactly one sentence — nothing else. Trigger when the user says "one line", "one sentence", "in a sentence", "一句話", "explain in one line", or asks what something is and clearly wants no more than that. Different from tldr: tldr gets 5 lines and can add key facts or a verdict. one-line gets one sentence and stops. Output: one bold sentence. No bullets, no heading, no follow-up offer.

  • 2 stars
  • 0 votes
  • 0 copies
  • 0 views
  • Added October 19, 2026
ai-agentsgoapidatabase

Works with

  • cli
  • api

Security analysis

A100/100

Scanned October 19, 2026

npx -y skills add Jsnnmsc/10000x-engineer --skill one-line --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of One Line?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for One Line
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/jsnnmsc-one-line/badge)](https://www.skillsdirectory.com/skills/jsnnmsc-one-line)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: one-line
description: |
  Explain it in exactly one sentence — nothing else.

  Trigger when the user says "one line", "one sentence", "in a sentence", "一句話", "explain in one line", or asks what something is and clearly wants no more than that.

  Different from tldr: tldr gets 5 lines and can add key facts or a verdict. one-line gets one sentence and stops.

  Output: one bold sentence. No bullets, no heading, no follow-up offer.
disable-model-invocation: true
---

# One Line

Goal: user reads one sentence and has the right mental model.

## Rules

- Respond in the same language the user wrote in. If they write in Chinese, reply in Chinese. Technical terms (JWT, Redis, API…) stay in English.
- **Max 1 sentence. Hard limit.** No second sentence, no bullets, no heading, no code block, no "let me know if you want more".
- The sentence is **bold** — wrap the whole thing in `**`, punctuation included. Nothing sits outside the bold.
- Under 25 words. If it doesn't fit, you haven't found the core yet — cut the qualifiers, not the meaning.
- No section headers and no closing verdict line — the sentence *is* the verdict.
- Say what it *does* or *is for*, not what category it belongs to. "A tamper-proof token that proves who you are" beats "an open standard for representing claims".
- No analogies unless the analogy *is* the shortest true explanation.
- No hedging ("generally", "basically", "essentially", "kind of"). Commit.
- No emoji.

## Output format

```
**[One sentence.]**
```

That's the whole output.

## Examples

`/one-line JWT`
```
**A signed token the client carries so the server can verify who they are without storing a session.**
```

`/one-line 為什麼要用 Redis`
```
**把讀取頻繁但變動不多的資料放在記憶體裡,避免每次都打 database。**
```

`/one-line this function`
```
**Retries the payment call with exponential backoff, giving up after three failures.**
```

## Anti-examples

Don't write "JWT is a JSON Web Token. It's used for auth." — that's two sentences and the
first one says nothing.

Don't write "A signed token for auth — see the docs for details." Pointing elsewhere is not
an explanation.

If no request was supplied, ask: "One line on what?"

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.