short-answers
Skill0 starsChanged 58 days agoMITView on GitHub
Keep replies as short as the job allows without dropping what the user needs to act on. Use when the user says answers are too long, asks for concise, terse, caveman-style or ADHD-friendly replies, mentions output tokens or verbosity, or wants lean reply defaults in AGENTS.md or CLAUDE.md.
1 other project uses a version of this.
Adopted by
Most recently taken up by
- diflowrin/plain-modetheir own version, first seen Sep 13, 2026
Other versions1 project uses a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 1
- diflowrin/plain-mode★ 0same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository10 other files it gives its agents.
- cheap-subagentsSkill
- context-auditSkill
- limit-window-warmupSkill
- pinpointSkill
- progress-ledgerSkill
- quiet-toolsSkill
- skillifySkill
- token-saverSkill
- usage-budgetSkill
- workflow-optimizerSkill
The file
We split this skill into the rules it gives an agent
2 rules. About 0 are common, 2 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈2
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 18
No closing recap, no "let me know if…", no menu of optional extras.
- Distinctive: line 20
Keep exact values, commands, paths and error text.
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Short answers
Output tokens are the priciest tokens: on current Claude models output costs five times as much as input, and other vendors price it in the same direction. A token is about three quarters of an English word, so a 100-word reply is roughly 130 tokens (about 30% more on newer Claude tokenizers). In agent sessions, though, output is usually the smaller part of the bill (the context re-read on every turn is the big one), so this skill trims real waste but doesn't replace the context habits in quiet-tools and context-audit.
Rules
- Lead with the result or the answer. No preamble, no restating the request, no "I'll now…".
- For a code change: what changed and where (
file:line), plus one line of why if it isn't obvious. Don't narrate the diff. - No closing recap, no "let me know if…", no menu of optional extras.
- Lists only when they save words; tables only when comparing.
- Keep exact values, commands, paths and error text. Short never means vague.
- Expand when asked, when the user has to decide something, or when there's a risk they need to know about.
- Plain words over shorthand. Caveman-style grammar ("why use many token") trims a little more, but an independent test on real agent work measured only single-digit savings, because code and tool calls dominate the output. Use it only if the user asks for it.
Reasoning is output too
Hidden reasoning is billed as output. For mechanical work, a lower reasoning effort saves more than trimming the visible reply:
- Claude Code:
/effort low(up tomax) for the session. - Codex: the effort picker in
/model, ormodel_reasoning_effortin~/.codex/config.toml. - OpenAI API:
reasoning.effort, plustext.verbosity(low,medium,high) for the length of the visible answer.
Changing effort or model partway through a session can throw away the prompt cache on some hosts (Claude Code documents this), so pick it at the start.
Make it permanent
One line for the instruction file (AGENTS.md, CLAUDE.md or GEMINI.md):
Keep responses as short as the task allows: lead with the result, skip preamble, recaps and closing offers, and expand only when asked.Claude Code also ships a built-in style for this: /output-style concise. It leads with the result, skips narration, and still keeps error reports, security warnings and confirmations for destructive actions complete.
Codex already defaults its current models to low verbosity. model_verbosity = "low" in ~/.codex/config.toml sets it explicitly (it applies to Responses-API providers).
Before you finish
- The first sentence answers the question or states the result.
- No sentence only restates, previews or offers.
- Every command, path and error the user needs is still there.
Part of agent-skills · MIT · improvements welcome as issues.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

