Make this reply brief — the answer only, sized to the question. Use when the user writes /briefly anywhere in a prompt, or explicitly asks to keep it brief or for a TL;DR.
Scanned 10/7/2026
npx -y skills add dgilford/ai-science-toolkit --skill briefly --agent claude-codeInstalls into .claude/skills of the current project.
Are you the author of Briefly?
Add the live security badge to your README — it updates automatically with every re-scan.
[](https://www.skillsdirectory.com/skills/dgilford-briefly)More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.
---
name: briefly
description: Make this reply brief — the answer only, sized to the question. Use when the user writes /briefly anywhere in a prompt, or explicitly asks to keep it brief or for a TL;DR.
argument-hint: "[question] [length, e.g. '1 line' | '50 words']"
catalog:
order: 75
summary: 'Cap the length and detail of a reply — the answer only, sized to the question. Works mid-prompt.'
---
Answer **briefly**, for this reply only.
**The question** is everything the user wrote, minus `/briefly` itself — don't trust command arguments to hold it (stacked with another skill, `/briefly` can receive that skill's arguments). An explicit length ("1 line", "50 words", "3 bullets") overrides the defaults below.
**With another skill:** wherever `/briefly` appears — before, after, or inside another command's arguments — it caps only your reply's prose. Run the other skill's steps in full, and never shorten content that skill writes (log entries, files, commits); a literal `/briefly` in such content stays verbatim.
## Default budget (pick by question type)
| Question | Budget |
|---|---|
| Fact, yes/no, lookup | 1–2 sentences. Yes/no leads with the word. |
| "How do I…" / explain | ≤ 5 bullets or ~80 words |
| Choice / design / compare | Recommendation + ≤ 3 reasons. No option survey. |
| Code | Minimal snippet, ≤ 1 line of prose around it |
| Review / debug finding | Top 1–3 issues, most severe first |
The budget is a ceiling, not a target. When unsure, take the smaller one.
## Cut
- Preamble, restating the question, closing recaps, offers of further help.
- Headers and sections — plain sentences or a short list instead (this overrides house-style section headers).
- Secondary caveats; background the user already has.
- Exploration beyond what the answer needs — fewer tool calls, not just fewer words.
## Keep
- **At most one caveat** — the one that would change the answer or its interpretation (a baseline, a sign convention, a NaN/masking trap). One clause, not a paragraph. A second caveat goes in the closing "Skipped:" line, not the body.
- Uncertainty when present: "unsure; likely X".
- **Length is capped, rigor isn't:** verify what the answer depends on.
If something material was cut, end with **one** line naming it (e.g. "Skipped: the 2-D case.").
Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.
No comments yet. Be the first to comment!