AI Skill Report Card

Building Claude Skills

Creates and manages Claude MCP skills following best practices from the awesome-claude-skills repository. Use when developing MCP tools, creating skill repositories, or contributing to Claude ecosystems.

A-85·Jun 14, 2026·Source: Web
15 / 15
Bash
# Clone skill repository template git clone https://github.com/your-username/claude-skills cd claude-skills # Create new skill file touch skills/analyzing-data.md # Add YAML frontmatter and content echo '--- name: analyzing-data description: Analyzes datasets and generates insights with visualizations. Use when exploring data patterns or creating reports. ---' > skills/analyzing-data.md
Recommendation
Reduce verbosity in the Best Practices section - Claude already understands naming conventions and file organization basics
15 / 15

Progress:

  • Research existing skills - Check awesome-claude-skills for similar tools
  • Define skill scope - Single responsibility, clear trigger phrases
  • Create YAML frontmatter - Gerund name, third-person description
  • Write quick start - Immediate actionable example
  • Document workflow - Step-by-step process with checkboxes
  • Add concrete examples - Input/output pairs
  • List best practices - Guidelines and tips
  • Note common pitfalls - What to avoid
  • Test skill - Verify with Claude MCP server
  • Submit to repository - Create PR with proper documentation
Recommendation
Consolidate the multiple checklists (Progress, Quality Checklist) into one comprehensive workflow checklist
18 / 20

Example 1: Data Analysis Skill

YAML
--- name: analyzing-sales-data description: Processes sales datasets to identify trends, seasonal patterns, and revenue insights. Use when examining business performance or forecasting. ---

Example 2: Code Review Skill

YAML
--- name: reviewing-python-code description: Conducts thorough Python code reviews focusing on performance, security, and maintainability. Use when auditing codebases or mentoring developers. ---

Example 3: Content Creation Skill

YAML
--- name: writing-technical-docs description: Creates comprehensive technical documentation with proper structure and examples. Use when documenting APIs, software architecture, or user guides. ---
Recommendation
Add more concrete input/output examples showing the actual skill content structure, not just YAML frontmatter

Naming Convention:

  • Use gerund form: processing-images, managing-databases
  • Kebab-case only: lowercase, hyphens, no spaces
  • Be specific: analyzing-financial-data vs analyzing-data

Description Guidelines:

  • Start with action verb in third person
  • Include specific use cases and trigger phrases
  • Keep under 1024 characters
  • Mention domain/context when relevant

Content Structure:

  • Lead with executable code or immediate steps
  • Use progress checklists for complex workflows
  • Provide concrete input/output examples
  • Include error handling and edge cases

Repository Organization:

skills/
├── data-analysis/
│   ├── analyzing-datasets.md
│   └── visualizing-metrics.md
├── development/
│   ├── reviewing-code.md
│   └── debugging-applications.md
└── content/
    ├── writing-documentation.md
    └── creating-presentations.md

Avoid These Mistakes:

  • Generic names: helper-tool vs parsing-json-apis
  • First/second person: "I analyze" or "You can" vs "Analyzes"
  • Missing triggers: Description without use cases
  • Over-explanation: Defining basic concepts Claude knows
  • Multiple responsibilities: One skill = one clear purpose
  • Vague examples: Abstract descriptions vs concrete input/output
  • No error handling: Ignoring edge cases and failures

Quality Checklist:

  • Name follows gerund kebab-case format
  • Description is third-person with clear triggers
  • Quick start provides immediate value
  • Examples show concrete input/output
  • Workflow has actionable steps
  • Best practices are specific and useful
  • Common pitfalls prevent real mistakes

Integration Tips:

  • Test skills with Claude MCP server before submitting
  • Follow repository contribution guidelines
  • Include proper README and documentation
  • Tag skills appropriately for discoverability
  • Consider skill dependencies and prerequisites
0
Grade A-AI Skill Framework
Scorecard
Criteria Breakdown
Quick Start
15/15
Workflow
15/15
Examples
18/20
Completeness
20/20
Format
15/15
Conciseness
12/15