agentleFS
yourkenikeyourkenike/ciana-parrot/skills/skill-creation/SKILL.md

skill-creation

Skill1 starsChanged 9 months agoMITView on GitHub

Guide for creating, updating, and managing skills. Use when: you need to extend your capabilities with a new skill, or the user asks you to create one. Skills are persistent Markdown instruction modules auto-discovered at runtime.

  • Reads credentials

14 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

9 rules. 0 are common, 7 are distinctive, and 2 are very specific to this project.

  • 0

    Common

    Most of the 11 edited versions we read keep these.

  • 7

    Distinctive

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Create the file using write_file:

    line 26
  2. Distinctive:

    Always tell the user when you create a new skill

    line 33
  3. Distinctive:

    Every skill needs a YAML frontmatter block followed by Markdown instructions:

    line 37
  4. Distinctive:

    Follow this structure for the Markdown body:

    line 95
  5. Distinctive:

    Be procedural: write steps the agent can follow, not conceptual explanations

    line 122
  6. Very specific:

    Only create SKILL.md files — never create Python scripts or executable code in skill directories

    line 130
  7. Distinctive:

    Always verify the skill works by reading it back after creation

    line 135
  8. Very specific:

    Update: use edit_file or write_file to modify skills/<name>/SKILL.md

    line 139
  9. Distinctive:

    Delete: remove the skill directory (tell the user first)

    line 140

“Common” is measured: kept by at least 6 in 10 of the 11 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Skill Creation

You can create new skills to extend your own capabilities. Skills are Markdown instruction modules that persist across conversations and require no restart.

When to Create a Skill

  • A recurring task that benefits from a structured workflow
  • Domain knowledge worth preserving across conversations
  • A multi-step process the user asks about repeatedly
  • Integration with a new CLI tool or API

When NOT to Create a Skill

  • One-off tasks that won't recur
  • Information that belongs in MEMORY.md (user preferences, personal facts)
  • Temporary instructions or experiment notes

How to Create a Skill

  1. Choose a name: lowercase, digits, hyphens only (e.g., meeting-notes, code-review)
  2. Create the file using write_file:
write_file("skills/<name>/SKILL.md", content)
  1. The skill is available immediately on the next message — no restart needed
  2. Always tell the user when you create a new skill

SKILL.md Format

Every skill needs a YAML frontmatter block followed by Markdown instructions:

---
name: <skill-name>
description: "Brief description of what this skill does and when to use it. Max 1024 chars."
---

# Skill Title

Instructions in markdown...

Frontmatter Fields

FieldRequiredDescription
nameYesMust match the directory name exactly
descriptionYesWhat the skill does and when to use it. Max 1024 chars
requires_envNoList of environment variables needed (skill is hidden if missing)
requires_bridgeNoHost gateway bridge name required (skill is hidden if bridge unavailable)
homepageNoExternal URL for reference documentation
metadataNoProvider-specific metadata (e.g., OpenClaw)

Frontmatter Examples

Simple skill (no dependencies):

---
name: meeting-notes
description: "Generate structured meeting notes from conversations. Use when the user asks to summarize or document a meeting."
---

Skill requiring an API key:

---
name: notion
description: "Manage Notion workspace: create pages, query databases, add content blocks. Requires NOTION_API_KEY."
requires_env:
  - NOTION_API_KEY
---

Skill requiring a host bridge:

---
name: spotify
description: "Control Spotify playback: play, pause, skip, search tracks/albums/playlists via spogo CLI."
requires_bridge: "spotify"
---

Body Structure

Follow this structure for the Markdown body:

# Skill Title

Brief one-line summary.

## When to Use

- Trigger phrase or scenario 1
- Trigger phrase or scenario 2

## When NOT to Use

- Scenario where another approach is better

## Commands / Workflow

Step-by-step instructions, code blocks, curl examples, etc.

## Notes

- Caveats, rate limits, known issues

Guidelines for the Body

  • Be procedural: write steps the agent can follow, not conceptual explanations
  • Include concrete examples: real commands, API calls, curl snippets
  • Use code blocks with language hints for syntax highlighting
  • Document known issues and workarounds — these save the most time
  • Keep it concise: aim for what's needed to execute, not exhaustive documentation

Rules

  • Only create SKILL.md files — never create Python scripts or executable code in skill directories
  • The name in frontmatter MUST match the directory name
  • One skill per directory, one SKILL.md per skill
  • You can add supporting files (configs, docs) in the same directory if needed
  • You can update or delete your own skills when they become outdated
  • Always verify the skill works by reading it back after creation

Updating & Deleting Skills

  • Update: use edit_file or write_file to modify skills/<name>/SKILL.md
  • Delete: remove the skill directory (tell the user first)
  • Changes take effect on the next message — no restart needed

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.