Skip to content
Back to search

llm-system-one-skill

System One (SYS1): a concise, opt-in AI skill with English and Traditional Chinese editions.

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/ncbajack0609--llm-system-one-skill/badge.svg)](/listings/ncbajack0609--llm-system-one-skill)

HTML

<a href="/listings/ncbajack0609--llm-system-one-skill"><img src="/api/listings/ncbajack0609--llm-system-one-skill/badge.svg" alt="findsafeskills safety rating" /></a>

Install

i

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

README.md · README · 36 lines
  1. # System One (SYS1)
  2. [繁體中文](README.zh-TW.md)
  3. System One is a short, opt-in instruction set for getting a direct answer with less searching, fewer tool calls, and less text. It uses the [Agent Skills format](https://agentskills.io/specification) and requires no particular agent, model, tool, or notes app. It does not change model settings, enforce a tool budget, or guarantee faster replies.
  4. ## Why System One?
  5. This skill takes its name and inspiration from Daniel Kahneman’s [*Thinking, Fast and Slow*](https://academic.macmillan.com/academictrade/9780374275631/thinkingfastandslow/), which contrasts fast, intuitive thinking (System 1) with slower, more deliberate thinking (System 2). It borrows that distinction as a practical rule for working with an LLM: answer directly when you can, and look deeper when the question calls for it.
  6. With a large notes collection—say, an Obsidian vault with more than 5,000 files—even a simple question can send the agent searching far more files than it needs. System One aims to cut that wait: use what’s already available first, search narrowly if needed, and stop once there’s enough to answer.
  7. ## Use it
  8. Pick **one** language edition. Each recognizes the same three triggers.
  9. - [English `SKILL.md`](skills/system-one-en/SKILL.md)
  10. - [繁體中文 `SKILL.md`](skills/system-one/SKILL.md)
  11. - **If your agent supports Agent Skills:** Follow your client's instructions to import the chosen folder or place it in a supported skills directory, such as `.agents/skills/` if your client scans it. Keep the folder name and `SKILL.md` filename. If the prefix alone does not load the skill, select it manually if your client offers that option, or use the prompt method below.
  12. - **Use it as a prompt without installing:** Copy the Markdown body below the YAML header into the chat and ask the agent to follow it. In your next message, start your request with a prefix below.
  13. For each request that should use System One, start with `SYS1`, `System One`, or `系統一`, for example:
  14. > SYS1: What should I check first if an app opens to a blank screen?
  15. ## What changes
  16. - Use available tools and other skills when needed, within the agent's permissions and required checks.
  17. - If you've designated a notes folder or knowledge base, the agent reads at most 3 distinct documents from it per task; snippets count. The limit does not apply to files outside it. No notes setup is required.
  18. - Replies in at most 700 characters by default, leading with the main point and using 1–2 sentences per extra point.
  19. - Ends with a self-rating such as `Confidence: 4.0/5`. A rating of 5 means high confidence; below 2.5, it suggests switching to normal mode for a thorough file search. The rating is a judgment, not a calibrated probability.
  20. ## License
  21. [MIT](LICENSE)

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.