A centralized repository for custom Claude Skills that can be imported into Claude Code or Claude Desktop.
Claude Skills are modular capabilities that extend Claude's functionality through organized directories containing instructions, scripts, and resources. They teach Claude how to complete specific repeatable tasks like creating documents with brand guidelines, analyzing data using organizational workflows, or automating personal tasks.
For project-level access:
cp -r skills/skill-name .claude/skills/For personal/global access:
cp -r skills/skill-name ~/.claude/skills/Skills are automatically discovered on the next session.
-
Create a
.zipfile of the skill directory:cd skills/skill-name zip -r skill-name.zip .
-
Open Claude Desktop → Settings → Capabilities → Skills
-
Click "Upload skill" and select the .zip file
Comprehensive expertise for Azure Static Web Apps including architecture, configuration, API integration with Azure Functions, authentication, routing, deployment, and CI/CD.
Use when:
- Building new Azure Static Web Apps
- Configuring SWA with React, Angular, Vue, Blazor, or other frameworks
- Integrating Azure Functions as serverless APIs
- Setting up authentication and authorization
- Configuring custom domains and SSL
- Troubleshooting deployment or runtime issues
- Implementing CI/CD with GitHub Actions
Location: skills/azure-swa/
Features:
- Complete architecture guidance
- Configuration examples (staticwebapp.config.json)
- API integration patterns
- Authentication and authorization setup
- Deployment and CI/CD workflows
- Security and performance best practices
- Comprehensive troubleshooting guide
Adds a complete blog feature to an existing Blazor WebAssembly Static Web App with Azure Functions backend and Azure File Share for markdown storage.
Use when:
- Implementing blog functionality in .NET Blazor WASM projects
- Integrating Azure Storage for content management
- Building content-driven Blazor applications
Location: skills/blazor-blog-feature/
Interacts with the FreeAgent accounting API to manage invoices, contacts, projects, expenses, timeslips, and other financial data.
Use when:
- Building integrations with FreeAgent accounting system
- Automating financial workflows
- Retrieving or analyzing FreeAgent data
Location: skills/freeagent-api/
Formats markdown files according to best practices and common style guidelines.
Use when:
- Cleaning up existing markdown documentation
- Standardizing README files across projects
- Ensuring consistent formatting in documentation
- Preparing markdown for publication
Location: skills/markdown-formatter/
Features:
- Standardizes headers, lists, and code blocks
- Fixes spacing and indentation
- Ensures consistent emphasis markers
- Validates link text and alt text
- Includes validation script
Comprehensive skill for working with Microsoft Graph API across all services including users, groups, mail, calendar, files (OneDrive/SharePoint), Teams, security, applications, and more.
Use when:
- Implementing Microsoft Graph API integrations
- Querying Microsoft 365 data
- Building applications that interact with Azure AD and Microsoft services
Location: skills/microsoft-graph/
claude-skills/
├── README.md # This file - getting started guide
├── CLAUDE.md # Repository documentation for Claude
├── skills/ # All skills stored here
│ └── skill-name/ # Individual skill directory
│ ├── SKILL.md # Required: skill entry point
│ ├── resources/ # Optional: supporting files
│ ├── templates/ # Optional: forms or structured prompts
│ └── scripts/ # Optional: utility scripts
└── .claude/ # Claude Code configuration
└── hooks/ # Repository hooks
mkdir -p skills/my-skill/resources
mkdir -p skills/my-skill/scripts
touch skills/my-skill/SKILL.mdEvery skill must have a SKILL.md file with YAML frontmatter:
---
name: my-skill
description: Brief description of what the skill does and when to use it
allowed-tools: Read, Edit, Bash # Optional: restrict tools
version: 1.0.0
---
# My Skill
Main skill instructions here...
## When to Use This Skill
- Describe use cases
## How to Use This Skill
1. Step-by-step instructions
2. Reference resources as needed
## Guidelines
- Key points
- Best practicesresources/style-guide.md- Detailed reference documentationresources/examples.md- Before/after examplesresources/checklist.txt- Structured checklistsscripts/helper.sh- Executable utilities
# Copy to personal skills directory
cp -r skills/my-skill ~/.claude/skills/
# Test with Claude Code
# Request a task that should trigger the skillgit add skills/my-skill/
git commit -m "Add my-skill for [purpose]"
git push- Keep
SKILL.mdunder 500 lines - Use progressive disclosure (reference additional files)
- Include clear examples
- Provide structured checklists for complex workflows
- Name: lowercase, hyphenated, max 64 characters
- Description: third-person, clear purpose, max 1024 characters
- Version: semantic versioning (1.0.0)
- Write in third person for descriptions
- Use second person for instructions
- Be specific and actionable
- Include examples of expected inputs/outputs
- Only name/description (~100 tokens) is pre-loaded
- Full SKILL.md loaded when Claude determines relevance
- Additional resources loaded only when referenced
- Use scripts for computationally-intensive operations
- Test with representative tasks
- Verify skill triggers appropriately
- Check that referenced files are accessible
- Validate scripts execute correctly
You can restrict which tools Claude can use within a skill by setting allowed-tools in the frontmatter:
---
name: safe-reader
description: Read files without modification capabilities
allowed-tools: Read, Grep, Glob
---Common tool combinations:
- Read-only:
Read, Grep, Glob - File editing:
Read, Edit, Write - Full access:
Read, Edit, Write, Bash, Grep, Glob
Structure skills in layers:
- Metadata: Name and description (always loaded)
- Core: Main SKILL.md body (loaded when relevant)
- Details: Referenced files (loaded only when needed)
Include scripts that Claude can run:
- Validation scripts (check before processing)
- Helper utilities (complex operations)
- Test scripts (verify correctness)
Example:
## Validation
Before formatting, run the validation script:
```bash
./skills/markdown-formatter/scripts/validate-markdown.sh file.md
### Hooks Integration
Skills can work with Claude Code hooks:
- **SessionStart**: Load skill context on startup
- **PreToolUse**: Validate operations before execution
- **Stop**: Ensure work completeness
## Contributing
1. Fork this repository
2. Create your skill in `skills/your-skill-name/`
3. Test thoroughly
4. Submit a pull request with:
- Clear description of skill purpose
- Example use cases
- Testing results
## Examples and Templates
### Minimal Skill
```markdown
---
name: simple-skill
description: Does one specific thing well
version: 1.0.0
---
# Simple Skill
This skill helps you do X.
## Instructions
1. Do this
2. Then this
3. Finally this
## Output
Provide results in this format:
- Summary
- Details
See skills/markdown-formatter/ for an example of a comprehensive skill with:
- Detailed SKILL.md
- Multiple resource files
- Executable validation script
- Structured checklists
- Verify
SKILL.mdexists with valid frontmatter - Check YAML formatting (use
---delimiters) - Ensure
namematches directory name - Confirm proper file permissions
- Make description more specific about when to use
- Test with explicit mention of skill name
- Check that task matches described use cases
- Reduce SKILL.md size (keep under 500 lines)
- Move detailed content to referenced files
- Use scripts for heavy operations
- Claude Skills Documentation
- Skill Best Practices
- Anthropic Skills Repository
- Awesome Claude Skills
- CLAUDE.md - Detailed repository documentation
This repository is provided as-is for organizing and sharing Claude Skills. Individual skills may have their own licenses.
For issues or questions:
- Check the troubleshooting section
- Review CLAUDE.md for detailed documentation
- Consult official Claude documentation
- Open an issue in this repository