Agent Skills

Skills

Author

Tony D

Published

February 23, 2026

What is Skills

Agent Skills are a lightweight, open format for extending AI agent capabilities with specialized knowledge and workflows.

At its core, a skill is a folder containing a SKILL.md file. This file includes metadata (name and description, at minimum) and instructions that tell an agent how to perform a specific task. Skills can also bundle scripts, templates, and reference materials.

my-skill/

├── SKILL.md # Required: instructions + metadata

├── scripts/ # Optional: executable code

├── references/ # Optional: documentation

└── assets/ # Optional: templates, resources

Skill.md example

---
name: code-review
description: Code review and quality inspection expert
allowed-tools: Read, Grep, Bash, Edit
---

# Code Review Expert

You are an experienced code reviewer who follows industry best practices.

## Review Focus
1. **Code Quality**
   - Naming conventions
   - Code complexity
   - Code duplication

2. **Security**
   - SQL injection risks
   - XSS vulnerabilities
   - Authentication and authorization issues

3. **Performance**
   - Algorithm efficiency
   - Resource usage
   - Caching strategies

adding skills

install

Code

Terminal

npx skills add vercel-labs/agent-skills

Code

Terminal

# List skills in a repository
npx skills add vercel-labs/agent-skills --list

Code

Terminal

# Full GitHub URL
npx skills add https://github.com/vercel-labs/skills/tree/main/skills/find-skills

Search for skills

Search for PPT related skills

#| eval: false
#| filename: "Terminal"
npx skills find xxxx

Check for skill updates

Code

Terminal

npx skills check

Update all installed skills

Code

Terminal

npx skills update

How Agents Use Skills

Agents don’t load all skills into memory at once. Instead, they follow a “Just-In-Time” activation pattern:

Detection: The agent scans the description of all available skills.
Matching: When a user’s request matches a skill’s description, the agent reads the full SKILL.md content.
Execution: The agent follows the instructions in the SKILL.md, potentially calling scripts or using references bundled in the skill folder.

set up for different Agents

user level

~/.claude/skills/<skill-name>/SKILL.md

project level

your_project/.claude/skills/<skill-name>/SKILL.md

user level

~/.agents/skills/<skill-name>/SKILL.md

project level

your_project/.agent/skills/<skill-name>/SKILL.md

user level

~/.opencode/skill/<skill-name>/SKILL.md

project level

your_project/.agent/skills/<skill-name>/SKILL.md

user level

~/.gemini/antigravity/skills/<skill-name>/SKILL.md

project level

your_project/.agent/skills/<skill-name>/SKILL.md

user level

~/.trae/skills/<skill-name>/SKILL.md

project level

your_project/.trae/skills/<skill-name>/SKILL.md

user level

~/.codebuddy/skills/<skill-name>/SKILL.md

project level

your_project/.codebuddy/skills/<skill-name>/SKILL.md

--- title: "Agent Skills" author: "Tony D" date: "2026-02-23" categories: [AI, Skills] image: "images/my screenshots.png" format: html: code-fold: show code-tools: true code-copy: false execute: warning: false editor: markdown: wrap: 72 --- # What is Skills Agent Skills are a lightweight, open format for extending AI agent capabilities with specialized knowledge and workflows. At its core, a skill is a folder containing a SKILL.md file. This file includes metadata (name and description, at minimum) and instructions that tell an agent how to perform a specific task. Skills can also bundle scripts, templates, and reference materials. my-skill/ ├── SKILL.md # Required: instructions + metadata ├── scripts/ # Optional: executable code ├── references/ # Optional: documentation └── assets/ # Optional: templates, resources # Skill.md example ```markdown --- name: code-review description: Code review and quality inspection expert allowed-tools: Read, Grep, Bash, Edit --- # Code Review Expert You are an experienced code reviewer who follows industry best practices. ## Review Focus 1. **Code Quality** - Naming conventions - Code complexity - Code duplication 2. **Security** - SQL injection risks - XSS vulnerabilities - Authentication and authorization issues 3. **Performance** - Algorithm efficiency - Resource usage - Caching strategies ``` # adding skills ## install ```{python} #| eval: false #| filename: "Terminal" npx skills add vercel-labs/agent-skills ``` ```{python} #| eval: false #| filename: "Terminal" # List skills in a repository npx skills add vercel-labs/agent-skills --list ``` ```{python} #| eval: false #| filename: "Terminal" # Full GitHub URL npx skills add https://github.com/vercel-labs/skills/tree/main/skills/find-skills ``` ## Search for skills Search for PPT related skills ```{bash} #| eval: false #| filename: "Terminal" npx skills find xxxx ``` ## Check for skill updates ```{python} #| eval: false #| filename: "Terminal" npx skills check ``` ## Update all installed skills ```{python} #| eval: false #| filename: "Terminal" npx skills update ``` # How Agents Use Skills Agents don't load all skills into memory at once. Instead, they follow a "Just-In-Time" activation pattern: 1. **Detection**: The agent scans the `description` of all available skills. 2. **Matching**: When a user's request matches a skill's description, the agent reads the full `SKILL.md` content. 3. **Execution**: The agent follows the instructions in the `SKILL.md`, potentially calling scripts or using references bundled in the skill folder. # set up for different Agents ::: panel-tabset ## Claude Code CLI ### user level `~/.claude/skills/<skill-name>/SKILL.md` ### project level `your_project/.claude/skills/<skill-name>/SKILL.md` ## Gemini CLI ### user level `~/.agents/skills/<skill-name>/SKILL.md` ### project level `your_project/.agent/skills/<skill-name>/SKILL.md` ## Opencode CLI ### user level `~/.opencode/skill/<skill-name>/SKILL.md` ### project level `your_project/.agent/skills/<skill-name>/SKILL.md` ## Antigravity IDE ### user level `~/.gemini/antigravity/skills/<skill-name>/SKILL.md` ### project level `your_project/.agent/skills/<skill-name>/SKILL.md` ## Trae IDE ### user level `~/.trae/skills/<skill-name>/SKILL.md` ### project level `your_project/.trae/skills/<skill-name>/SKILL.md` ## CodeBuddy IDE ### user level `~/.codebuddy/skills/<skill-name>/SKILL.md` ### project level `your_project/.codebuddy/skills/<skill-name>/SKILL.md` :::