TopRank Skills Documentation

Introduction

TopRank Skills are modular capabilities that extend Claude's functionality. Built on GitHub repositories, skills enable Claude to perform specialized tasks, integrate with external services, and provide custom workflows tailored to your needs.

With over 80,000 community-contributed skills available, you can find solutions for everything from code analysis and data processing to API integrations and automation workflows.

Installation

To use TopRank Skills, you need to install Claude Code CLI. This command-line tool manages skill installation, configuration, and execution.

# Install via npm (recommended)
npm install -g @anthropic-ai/claude-code

# Or using yarn
yarn global add @anthropic-ai/claude-code

# Verify installation
claude-code --version

System Requirements

• Node.js: v18.0 or higher
• npm: v9.0 or higher (or yarn v3.0+)
• Git: v2.20 or higher (for cloning skill repositories)
• Operating System: macOS, Linux, or Windows (WSL2 recommended)
• Claude API Key: Available from your Claude account

Quick Start

Get started with TopRank Skills in three simple steps:

export CLAUDE_API_KEY="your-api-key-here"

claude-code skill add skill-creator

claude "Use the skill-creator skill to help me build a new skill"

What are Skills?

Skills are self-contained units of functionality that Claude can invoke during conversations. Each skill is defined by a JSON manifest that specifies:

• Name and Description: How Claude identifies and understands the skill
• Parameters: What inputs the skill accepts
• Tools: What actions the skill can perform
• Permissions: What resources the skill can access

Finding Skills

Browse our directory at https://toprankskills.com/zh/skill to discover skills by category:

• Development: Code analysis, testing, CI/CD automation
• Data: Processing, visualization, analysis
• Integration: API connectors, webhooks, third-party services
• Content: Writing, editing, translation
• Automation: Workflows, scheduled tasks, notifications

Installing Skills

Install skills using the CLI or by cloning from GitHub:

Via CLI (Recommended)

# Install by skill name
claude-code skill add skill-name

# Install from GitHub URL
claude-code skill add https://github.com/user/skill-repo

# Install specific version
claude-code skill add skill-name@1.2.0

Via GitHub Clone

git clone https://github.com/user/skill-repo.git
cd skill-repo
claude-code skill add .

Using Skills

Once installed, skills are automatically available in Claude conversations. Reference them naturally in your prompts:

Managing Skills

# List installed skills
claude-code skill list

# Update a skill
claude-code skill update skill-name

# Remove a skill
claude-code skill remove skill-name

# Show skill details
claude-code skill info skill-name

Skill Structure

A basic skill consists of:

my-skill/
├── skill.json          # Skill manifest (required)
├── README.md           # Documentation
├── .env.example        # Environment variables template
└── tools/              # Tool implementations
    └── main.js

Creating Your First Skill

Let's create a simple skill that searches GitHub repositories:

Step 1: Create Skill Manifest

{
  "name": "github-repo-search",
  "version": "1.0.0",
  "description": "Search GitHub repositories by keyword",
  "author": "Your Name",
  "license": "MIT",
  "parameters": {
    "type": "object",
    "properties": {
      "query": {
        "type": "string",
        "description": "Search query"
      },
      "language": {
        "type": "string",
        "description": "Programming language filter",
        "optional": true
      },
      "limit": {
        "type": "number",
        "description": "Maximum results",
        "default": 10
      }
    },
    "required": ["query"]
  },
  "tools": {
    "search": {
      "description": "Execute repository search",
      "handler": "./tools/search.js"
    }
  }
}

Step 2: Define Parameters

Parameters use JSON Schema validation. Supported types include:

• string - Text values
• number - Numeric values
• boolean - True/false
• array - Lists of values
• object - Nested structures

Step 3: Test Locally

claude-code skill test ./my-skill --params '{"query": "TopRank Skills", "limit": 5}'

Step 4: Publish to GitHub

git init
git add .
git commit -m "Initial skill creation"
git remote add origin https://github.com/yourusername/github-repo-search.git
git push -u origin main

Environment Variables

Store sensitive data like API keys in environment variables:

GITHUB_TOKEN=ghp_your_token_here
API_ENDPOINT=https://api.example.com
DEBUG=false

Access in your skill code using process.env.VARIABLE_NAME.

Custom Tools

Tools are JavaScript/TypeScript functions that skills can invoke:

export default async function search({ query, language, limit }) {
  const response = await fetch(
    `https://api.github.com/search/repositories?q=${query}${language ? `+language:${language}` : ''}&per_page=${limit}`,
    {
      headers: {
        'Authorization': `token ${process.env.GITHUB_TOKEN}`
      }
    }
  );

  const data = await response.json();
  return data.items.map(repo => ({
    name: repo.full_name,
    description: repo.description,
    stars: repo.stargazers_count,
    url: repo.html_url
  }));
}

Hooks System

Hooks allow skills to respond to events:

• onInstall - Runs when skill is installed
• onUpdate - Runs when skill is updated
• onUninstall - Runs before skill removal
• beforeInvoke - Runs before tool execution
• afterInvoke - Runs after tool execution

Skill Composition

Skills can depend on other skills, creating powerful workflows:

{
  "name": "advanced-analyzer",
  "dependencies": {
    "github-analyzer": "^1.0.0",
    "code-quality": "^2.1.0"
  }
}

CLI Commands Reference

Skill API Reference

Core API methods available in skill tools:

Configuration Options

Configure Claude Code behavior in ~/.claude/config.json:

{
  "apiKey": "your-api-key",
  "model": "claude-3-sonnet",
  "skillsDirectory": "~/.claude/skills",
  "maxTokens": 4096,
  "temperature": 0.7,
  "debug": false
}

Security Best Practices

• Never hardcode secrets: Use environment variables for API keys and tokens
• Validate inputs: Always validate and sanitize user inputs
• Use HTTPS: Ensure all API calls use secure connections
• Limit permissions: Request only the minimum permissions needed
• Review dependencies: Audit third-party packages before use
• Update regularly: Keep skills and dependencies up to date

Performance Optimization

• Cache responses: Store frequently accessed data
• Batch operations: Combine multiple API calls when possible
• Lazy loading: Load resources only when needed
• Use streaming: Stream large responses instead of buffering
• Set timeouts: Prevent hanging requests

Debugging Skills

Enable debug mode for verbose logging:

export DEBUG=claude:*
claude-code skill test ./my-skill --verbose

Troubleshooting

Common Issues

Frequently Asked Questions