2026 Claude Code Skills Template Collection (100+ Free Examples)

In 2026, AI coding assistants have evolved from chat completions to orchestratable Agent workflows. Claude Code Skills (Anthropic's terminal agent) and Cursor Agent Skills (built into the IDE) share the same SKILL.md spec — Markdown files that teach the AI how to handle specific scenarios.

If you're still pasting "please review this PR using our team standards" into every conversation, it's time to write that into a Skill once and reuse it forever. This guide provides 120+ ready-to-use Skill names and trigger scenarios, plus 8 copy-paste templates covering the full development pipeline from code review to DevOps.

What are Skills? How do they relate to Rules and MCP?

In short: Rules set the baseline, Skills teach the method, MCP connects tools.

Claude Code Skills typically live in ~/.claude/skills/ (personal) or .claude/skills/ (project). Cursor uses ~/.cursor/skills/ and .cursor/skills/. The format is interchangeable — all templates in this guide work in both.

SKILL.md standard structure

Each Skill is a directory with SKILL.md as the core file:

my-skill/
├── SKILL.md          # Required: main instructions + YAML frontmatter
├── reference.md      # Optional: detailed reference
├── examples.md       # Optional: input/output examples
└── scripts/          # Optional: executable scripts
    └── validate.sh

Minimal template

---
name: my-skill-name
description: Third-person description of WHAT it does and WHEN to trigger. Include keywords for discovery.
---

# My Skill Name

## Steps
1. First concrete action
2. Second concrete action

## Output format
Describe the expected output structure.

Required frontmatter fields

Best practices (2026)

1. Be concise — only add what the agent doesn't already know
2. description is key — state WHAT + WHEN with trigger keywords
3. Main file ≤500 lines — move details to reference.md
4. Actionable steps — numbered lists, not vague guidance
5. Never create skills in ~/.cursor/skills-cursor/ — reserved for Cursor built-ins

120+ Skill quick reference

Below are 120 ready-to-use Skill names across 12 categories with trigger scenarios and suggested description keywords. Copy a name, create the directory, and customize.

1. Code review & quality (10)

2. Git & version control (10)

3. Testing (10)

4. DevOps & CI/CD (10)

5. Security (10)

6. Documentation (10)

7. Database (10)

8. Frontend (10)

9. Backend & API (10)

10. AI / LLM integration (10)

11. Mobile & cross-platform (10)

12. Project management & collaboration (10)

8 copy-paste SKILL.md templates

Copy these into .cursor/skills/<name>/SKILL.md or ~/.claude/skills/<name>/SKILL.md and customize.

Template 1: Code review

---
name: code-review
description: Review code changes for quality, security, and best practices. Use when reviewing pull requests, diffs, or when the user asks for a code review.
---

# Code Review

## Checklist
1. Read the full diff before commenting
2. Check for security issues (injection, XSS, secrets)
3. Verify error handling on all external calls
4. Confirm tests cover new logic
5. Flag breaking API changes

## Output Format
- **Critical**: must fix before merge
- **Suggestion**: optional improvement
- **Praise**: well-done patterns worth keeping

Template 2: Commit message

---
name: commit-message
description: Generate descriptive commit messages following Conventional Commits. Use when the user asks to write commit messages or review staged changes.
---

# Commit Message Generator

## Format
`<type>(<scope>): <subject>`

Types: feat, fix, docs, style, refactor, test, chore

## Rules
- Subject ≤ 72 characters, imperative mood
- Body explains WHY, not WHAT
- Reference issue numbers when applicable

Templates 3–8

The remaining templates cover unit testing, Dockerfile optimization, security review, API documentation, database migrations, and MCP server development — same structure as Templates 1–2 with scenario-specific checklists. See the Chinese version for full copy-paste bodies, or adapt from the Skill names in the catalog above.

Installation & usage

Cursor

1. Create: .cursor/skills/my-skill/SKILL.md (project) or ~/.cursor/skills/my-skill/SKILL.md (personal)
2. Mention the scenario in Agent chat — the agent matches description and loads the Skill
3. Or invoke explicitly: /my-skill-name or @my-skill in the prompt

Claude Code

1. Create: ~/.claude/skills/my-skill/SKILL.md
2. Describe the task in the terminal — the agent auto-matches
3. Project Skills go in .claude/skills/ and can be committed to Git

Team sharing

• Commit .cursor/skills/ to version control
• List available Skills and triggers in README.md
• Add examples.md per Skill showing expected I/O
• Review description wording when new team members join

Your dev environment matters too

Skills are only as good as the hardware and OS they run on. iOS-related Skills (swiftui-view, app-store-submit) require macOS + Xcode; large monorepo code-review Skills need sufficient memory and disk I/O.

For a full cloud macOS dev environment to pair with Claude Code or Cursor Agent, Kvmzen Mac cloud rental offers dedicated Mac mini M4 instances on a monthly plan — compatible with every Skill template in this guide.

Frequently asked questions