Skip to content
Skills Library
AI & agents · Writing

write-a-skill

/write-a-skill

How to write a good skill (a SKILL.md how-to) for a task you or the owner will do again. Use when the owner asks you to make, save or improve a skill, or after a tricky job went well and is worth keeping.

Repository
sara ↗
Licence
MIT
Last repository change
12.11.2026

Assessment and security

90/ 100Class: good
Format25 / 25
Frontmatter with name and description, valid name, description up to 1,024 characters, name matching the folder.
Discoverability25 / 25
Description explains when to use the skill, has a reasonable length and uses third person.
Content25 / 35
At least 150 words and at most 500 lines, headings, step lists, examples or code.
Hygiene15 / 15
No placeholders such as TODO, no Windows paths, with a licence.
Riskno signals detected in the scan
Files in this package
1
Scripts
0
HTML comments
0
URLs in SKILL.md
0
Deprecation

No signals detected. Repository changed on 12.11.2026, without references to retired models or APIs.

Licence

MIT, open: use and modify while retaining author attribution.

What it does and when to use it

How to write a good skill (a SKILL.md how-to) for a task you or the owner will do again. Use when the owner asks you to make, save or improve a skill, or after a tricky job went well and is worth keeping.

Original author description

How to write a good skill (a SKILL.md how-to) for a task you or the owner will do again. Use when the owner asks you to make, save or improve a skill, or after a tricky job went well and is worth keeping.

How it works

  1. 1. Check it is worth a skill
  2. 2. Name and description
  3. 3. The steps (the body)
  4. 4. Save it
  5. 5. Keep it true
Original SKILL.md headings · 36 lines, 353 words
SKILL.md preview
---
name: write-a-skill
description: "How to write a good skill (a SKILL.md how-to) for a task you or the owner will do again. Use when the owner asks you to make, save or improve a skill, or after a tricky job went well and is worth keeping."
license: MIT
metadata:
  "sara-version": "1"
---

# Write a skill

A skill is a folder with a `SKILL.md`: a short frontmatter and the steps. You read the name and
description of every skill at the start of a conversation, and open the steps only when a request
fits, so the description decides whether the skill is ever used.

## 1. Check it is worth a skill
- Worth it: a task with several steps that will come back (a deploy, a release note, a review,
  a weekly routine), or something you had to figure out the hard way.
- Not worth it: one-step work, or a fact (that belongs in memory, not in a skill).
- Look first with `skill_find`: if a skill for it exists, improve that one with `skill_update`.

## 2. Name and description
- **name**: lower-case words joined by single hyphens, 64 characters at most, e.g.
  `deploy-the-weather-app`. It becomes the folder name.
- **description**: one or two sentences saying what it does **and when to use it**
  ("... Use when the owner asks to put the weather app live"). Use the words the owner would use.

## 3. The steps (the body)
- Numbered steps, each one action: the command in backticks, the file to open, the check to run.
- Say how to tell it worked (the test count, the page that loads) and what to do when it doesn't.
- Write down the pitfalls you actually hit, briefly.
- Keep it under about 200 lines. Long details go in a file beside it (`references/notes.md`), and
  the steps say when to read it. Scripts go in `scripts/` and always ask the owner before running.
- No secrets, passwords or keys in a skill, ever.

## 4. Save it
- Save with `skill_save` (name, description, body). The owner sees the SKILL.md on the approval
  card and can say no.
- Tell the owner in one line what the skill is for and that they can edit it in
  Settings → Abilities → Skills.

## 5. Keep it true
- When a step turns out wrong later, fix it with `skill_update` and say why in one line.

Versions

RepositoryScoreLicence
This page · farena-ai/sara90MIT
borghei/Claude-Skills100MIT
alirezarezvani/claude-skills100MIT
redpanda-data/ui-harness100MIT
RubenGlez/rubstack100MIT
holl4ndtv/skills100MIT
yamcodes/arkenv100MIT
mkurman/zorai100MIT
AmirAzy/HoNspot100MIT

Database snapshot: 13.11.2026 · package version 1f29e66b0e78 (SHA-256 digest of the entire package)

Downloads in the last 30 days: 0 · ranking

Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.

Similar skills

  • 100write-skill
    Create a new agent skill, or improve an existing one, inside this skills repo, following its housekeeping rules. Use when asked to "write a skill", "make this a skill", "save this as a skill", "turn what we just did into a skill", or "improve the X skill".
  • 100prompt
    Build a great prompt through a conversational interview. Use this skill whenever the user wants help writing, improving, or crafting a prompt — including phrases like "help me write a prompt", "I need a prompt for", "how do I prompt Claude to", "make me a prompt that", "I want…
  • 100skill-development
    This skill should be used when the user asks to create a new skill, repair an existing skill, improve trigger descriptions, reorganize skill structure, or make a Claude skill more reusable and internally consistent.
  • 100skill-creator
    Create a new skill that packages a workflow into a reusable SKILL.md. Use when the owner says "create a skill", "make a new skill", "turn this into a skill", or asks to build a reusable process. Also called as a sub-step by promote-skills to draft candidate bodies.
  • 100skill-writing
    Use when adding a new skill or expertise pack to the vault, fixing an existing skill, or reviewing and adapting a skill that came from outside. The main chat loads it.
  • 100skill-creator
    Guide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends an AI agent's capabilities with specialized knowledge, workflows, or tool integrations.

Feedback and reports

Your message goes to Bartosz Sokoliński. An email address is needed only for a reply. See the privacy policy.