make-skill
Learn how to create an Agent Skill. Use when you have a capability to share or want to package something reusable.
Best use case
make-skill is best used when you need a repeatable AI agent workflow instead of a one-off prompt.
Learn how to create an Agent Skill. Use when you have a capability to share or want to package something reusable.
Teams using make-skill should expect a more consistent output, faster repeated execution, less prompt rewriting.
When to use this skill
- You want a reusable workflow that can be run more than once with consistent structure.
When not to use this skill
- You only need a quick one-off answer and do not need a reusable workflow.
- You cannot install or maintain the underlying files, dependencies, or repository context.
Installation
Claude Code / Cursor / Codex
Manual Installation
- Download SKILL.md from GitHub
- Place it in
.claude/skills/make-skill/SKILL.mdinside your project - Restart your AI agent — it will auto-discover the skill
How make-skill Compares
| Feature / Agent | make-skill | Standard Approach |
|---|---|---|
| Platform Support | Not specified | Limited / Varies |
| Context Awareness | High | Baseline |
| Installation Complexity | Unknown | N/A |
Frequently Asked Questions
What does this skill do?
Learn how to create an Agent Skill. Use when you have a capability to share or want to package something reusable.
Where can I find the source code?
You can find the source code on GitHub using the link provided at the top of the page.
SKILL.md Source
# Make Skill
How to create an Agent Skill that other agents (and humans) can discover and use.
## When to Use
- You've built something useful and want to share it
- You want to package a repeatable workflow
- You're teaching an agent how to do something specific
## Quick Start
### Using Scripts (Recommended)
```bash
# 1. First time only: create your skills repo
python scripts/init.py my-skills --public
# 2. Create a new skill inside your repo
python scripts/create.py my-skill-name --output ./my-skills/ --author "You"
# 3. Edit SKILL.md and add your code to scripts/
# 4. Publish (commits and pushes to your repo)
python scripts/publish.py ./my-skills/my-skill-name
```
### Manual Setup
1. Copy `references/template/` to your new skill folder
2. Edit the SKILL.md frontmatter (name, description)
3. Replace the example script with your code
4. Push to GitHub with topic `agentskills`
Done. Your skill is now discoverable.
## The Format
A skill is a folder:
```
my-skill/
SKILL.md # Required: frontmatter + instructions
scripts/ # Optional: executable code
references/ # Optional: additional documentation
assets/ # Optional: templates, data files
```
### SKILL.md Structure
```markdown
---
name: my-skill
description: What it does. When to use it. Be specific.
license: MIT
metadata:
author: you
version: "1.0"
parent: github.com/original/skill # if derived from another
---
# My Skill
Instructions for the agent...
## Usage
How to run it...
## Examples
Show inputs and outputs...
```
### Required Fields
| Field | Rules |
|-------|-------|
| `name` | Lowercase, hyphens only, matches folder name, max 64 chars |
| `description` | What it does AND when to use it, max 1024 chars |
### Optional Fields
| Field | Purpose |
|-------|---------|
| `license` | How others can use it |
| `metadata` | Arbitrary key-value pairs (author, version, parent) |
| `compatibility` | Environment requirements |
## Writing Good Instructions
The body of SKILL.md is what agents read. Make it clear:
**Do:**
- Start with when to use this skill
- Give concrete usage examples
- Show expected inputs and outputs
- List requirements and dependencies
- Handle edge cases
**Don't:**
- Assume context the agent won't have
- Write walls of text (keep it scannable)
- Bury the important stuff
## Making It Discoverable
### Option A: GitHub Topic
Add topic `agentskills` to your repository. Done.
### Option B: Monorepo
Keep multiple skills in one repo:
```
my-tools/
skill-one/
SKILL.md
skill-two/
SKILL.md
```
Tag the repo with `agentskills`. The find-skill searches inside.
### Option C: Local Only
Keep it in `~/skills/` or any folder. Configure find-skill to search there.
### Option D: Anywhere With a URL
Gist, pastebin, raw file host. As long as it's fetchable.
## Tracking Lineage
If your skill improves or derives from another:
```yaml
metadata:
parent: github.com/user/original-skill
parent-hash: abc123
```
Optional. Honor system. Helps the ecosystem.
## GitHub Token Setup
The scripts require a GitHub Personal Access Token to create repos, push code, and add topics.
### Creating Your Token
1. Go to GitHub → Settings → Developer settings → Personal access tokens → **Tokens (classic)**
2. Click "Generate new token (classic)"
3. Give it a name (e.g., "agent-skills")
4. Select the **`repo`** scope (full control of private repositories)
5. Generate and copy the token
### Storing Your Token
Create a `.env` file in the project root:
```
GITHUB_TOKEN=ghp_xxxxxxxxxxxx
```
Or set it as an environment variable: `GITHUB_TOKEN` or `GH_TOKEN`
**Important:** The `repo` scope is required to:
- Create repositories (init.py)
- Push code (publish.py)
- Add the `agentskills` topic (makes your skills discoverable)
## Scripts
### init.py
Initialize a skills monorepo (first time only):
```bash
python scripts/init.py my-skills # Private repo
python scripts/init.py my-skills --public # Public repo
```
Creates a GitHub repo with `agentskills` topic. All your skills go here.
### create.py
Create a skill from the template:
```bash
python scripts/create.py my-skill --output ./my-skills/
python scripts/create.py my-skill --author "You" --description "Does X" -o ./my-skills/
```
### publish.py
Commit and push a skill to your repo:
```bash
python scripts/publish.py ./my-skills/my-skill
python scripts/publish.py ./my-skills/my-skill --message "Update skill"
```
Commits the skill folder, adds the `agentskills` topic (if missing), and pushes to GitHub.
## Reference
See `references/format.md` for the full agentskills.io specification.
See `references/template/` for a working example to copy.Related Skills
adynato-aimake
Integrate with aimake's AI-powered delivery pipeline via MCP. Covers connecting to aimake, using code/docs/kanban tools, understanding the card-based system, and leveraging AI capabilities. Use when building integrations with aimake or using its MCP tools.
makefile
Best practices for authoring GNU Make Makefiles Triggers on: **/Makefile, **/makefile, **/*.mk, **/GNUmakefile
make-automation
Automate Make (Integromat) tasks via Rube MCP (Composio): operations, enums, language and timezone lookups. Always search tools first for current schemas.
zapier-make-patterns
No-code automation democratizes workflow building. Zapier and Make (formerly Integromat) let non-developers automate business processes without writing code. But no-code doesn't mean no-complexity ...
make-agent
Create GitHub Copilot custom agent files for specialized workflows. Use when asked to "create agent", "make custom agent", "scaffold agent", or when building task-specific agents like test-specialist, implementation-planner, or security-reviewer. Generates .agent.md files with YAML frontmatter and behavioral instructions.
agent-33-skills-maker
Create Claude Desktop skills from descriptions, SOPs, or existing agents. Three modes - create from scratch, convert SOPs to skills, or migrate agents to skill format. Outputs ready-to-deploy .zip packages.
bgo
Automates the complete Blender build-go workflow, from building and packaging your extension/add-on to removing old versions, installing, enabling, and launching Blender for quick testing and iteration.
moai-lang-r
R 4.4+ best practices with testthat 3.2, lintr 3.2, and data analysis patterns.
moai-lang-python
Python 3.13+ development specialist covering FastAPI, Django, async patterns, data science, testing with pytest, and modern Python features. Use when developing Python APIs, web applications, data pipelines, or writing tests.
moai-icons-vector
Vector icon libraries ecosystem guide covering 10+ major libraries with 200K+ icons, including React Icons (35K+), Lucide (1000+), Tabler Icons (5900+), Iconify (200K+), Heroicons, Phosphor, and Radix Icons with implementation patterns, decision trees, and best practices.
moai-foundation-trust
Complete TRUST 4 principles guide covering Test First, Readable, Unified, Secured. Validation methods, enterprise quality gates, metrics, and November 2025 standards. Enterprise v4.0 with 50+ software quality standards references.
moai-foundation-memory
Persistent memory across sessions using MCP Memory Server for user preferences, project context, and learned patterns