Skill Tester

Validate agent skills against the Agent Skills Specification v1.0 with comprehensive testing and reporting.

When to Use This Skill

Use skill-tester when you need to:

• Validate a new skill before deployment
• Test an existing skill after modifications
• Ensure spec compliance
• Catch structural issues early
• Generate validation reports

Testing Process

Phase 1: Initial Checks

Locate the Skill

1. Identify the skill directory path
2. Verify the directory exists
3. Confirm SKILL.md file is present (case-sensitive)

Phase 2: Run Validation

Use the Validation Script

Check available options:

python scripts/test_skill.py --help

Run full validation:

python scripts/test_skill.py /path/to/skill-directory

Phase 3: Interpret Results

Validation Report Structure

The validator checks:

✓ PASSED - Requirement met

✓ Directory exists: skill-name
✓ SKILL.md file exists
✓ YAML frontmatter is valid
✓ Required field present: 'name'
✓ Required field present: 'description'
✓ Skill name format is valid: 'skill-name'
✓ Directory name matches YAML name: 'skill-name'
✓ Description explains when to use the skill
✓ SKILL.md body has 1234 words (<5,000)

⚠ WARNINGS - Should be addressed

⚠ Description is 250 chars (recommended ~200)
⚠ SKILL.md body has 5,200 words (recommended <5,000)
⚠ Description should explain WHEN to use the skill
⚠ Possible hardcoded credential detected

✗ FAILED - Must be fixed

✗ SKILL.md file not found (case-sensitive)
✗ Invalid YAML frontmatter: mapping values are not allowed here
✗ Missing required field in YAML: 'description'
✗ Skill name exceeds 64 characters: 72
✗ Skill name must be hyphen-case: 'Skill_Name'
✗ Directory name 'skill-name' does not match YAML name 'skillname'
✗ Description exceeds 1024 characters (1500 chars)
✗ Referenced file does not exist: scripts/missing.py

Phase 4: Fix Issues

Priority 1: Fix All Errors

Errors prevent the skill from working correctly:

• Missing or invalid SKILL.md
• Invalid YAML syntax
• Missing required fields
• Name format violations
• Directory/YAML name mismatches
• Missing referenced files

Priority 2: Address Warnings

Warnings indicate quality issues:

• Description too long or vague
• SKILL.md too verbose
• Missing "when to use" guidance
• Potential security issues

Phase 5: Re-test

After fixes:

python scripts/test_skill.py /path/to/skill-directory

Verify:

• All errors resolved
• Warnings addressed
• Clean validation report

Validation Checks

1. Directory Structure

Checks:

• Directory exists and is a directory
• SKILL.md file exists (case-sensitive)

Common Issues:

• Using skill.md instead of SKILL.md
• Missing SKILL.md file
• Incorrect directory path

2. YAML Frontmatter

Checks:

• YAML syntax is valid
• Starts and ends with ---
• Required fields present: name, description
• Field types correct (strings, arrays, objects)

Common Issues:

• Missing opening or closing ---
• Invalid YAML syntax (indentation, special characters)
• Typos in field names
• Missing required fields

3. Skill Name

Checks:

• Hyphen-case format (lowercase, hyphens only)
• Maximum 64 characters
• Only Unicode alphanumeric and hyphens
• Directory name matches YAML name field exactly

Valid Examples:

• code-analyzer
• api-doc-generator
• test-runner

Invalid Examples:

• Code_Analyzer (underscores, capitals)
• codeAnalyzer (camelCase)
• code analyzer (spaces)

4. Description Field

Checks:

• String type
• Maximum 1024 characters (API limit)
• Recommended ~200 characters
• Explains WHAT and WHEN

Good Description:

description: Analyze code complexity using cyclomatic complexity metrics. Use when assessing code maintainability or identifying refactoring candidates.

Poor Description:

description: This skill helps with code.

5. File References

Checks:

• All referenced files exist
• Paths are correct relative to skill root
• Checks markdown links and inline code that reference local files

Common Issues:

• Typos in file paths
• Case sensitivity mismatches
• Absolute paths instead of relative
• References to non-existent files

6. Content Quality

Checks:

• Word count under 5,000 (warning at 5,000+)
• No hardcoded credentials (simple pattern check)
• Description includes "when" indicators

Security Patterns Checked:

• password: "..."
• api_key: "..."
• secret: "..."
• token: "..."

Testing Scripts

test_skill.py

Purpose: Main validation script for skill testing

Usage:

python scripts/test_skill.py <skill-directory>

Exit Codes:

• 0: All checks passed
• 1: One or more checks failed

Output: Formatted validation report with passed, warnings, and failed checks

validate_yaml.py

Purpose: Validate YAML frontmatter only

Usage:

python scripts/validate_yaml.py <skill-directory>

Checks:

• YAML syntax
• Required fields
• Field types and constraints

Examples

Example 1: Valid Skill

Skill Structure:

brand-guidelines/
└── SKILL.md

SKILL.md:

---
name: brand-guidelines
description: Apply brand visual identity guidelines including colors, typography, and spacing. Use when creating branded materials or reviewing designs.
license: MIT
---
# Brand Guidelines

[Content follows...]

Validation Result:

✓ PASSED (9)
  ✓ Directory exists: brand-guidelines
  ✓ SKILL.md file exists
  ✓ YAML frontmatter is valid
  ✓ Required field present: 'name'
  ✓ Required field present: 'description'
  ✓ Skill name format is valid: 'brand-guidelines'
  ✓ Directory name matches YAML name: 'brand-guidelines'
  ✓ Description explains when to use the skill
  ✓ SKILL.md body has 245 words (<5,000)

RESULT: PASSED - Skill is valid

Example 2: Skill with Errors

Skill Structure:

Code_Analyzer/
└── SKILL.md

SKILL.md:

---
name: code-analyzer
description: Analyzes code.
---
# Code Analyzer

Run the analysis script from the scripts directory.

Validation Result:

✗ FAILED (2)
  ✗ Directory name 'Code_Analyzer' does not match YAML name 'code-analyzer'

⚠ WARNINGS (1)
  ⚠ Description should explain WHEN to use the skill

RESULT: FAILED - Fix errors before using this skill

Fixes Required:

1. Rename directory to code-analyzer
2. Improve description to explain when to use

Example 3: Skill with Warnings

SKILL.md:

---
name: data-processor
description: This skill provides comprehensive data processing capabilities including cleaning, transformation, validation, analysis, and reporting across multiple data formats and sources with support for batch and streaming operations.
---

Validation Result:

✓ PASSED (7)
  [... passed checks ...]

⚠ WARNINGS (2)
  ⚠ Description is 210 chars (recommended ~200)
  ⚠ Description should explain WHEN to use the skill

RESULT: PASSED with warnings - Consider addressing warnings

Improvements:

description: Process, clean, and transform data across multiple formats. Use when preparing datasets for analysis or integrating data from different sources.

Common Pitfalls

Name Mismatches

❌ Wrong:

Directory: data-analyzer
YAML: data_analyzer

✓ Correct:

Directory: data-analyzer
YAML: data-analyzer

Case Sensitivity

❌ Wrong:

skill.md
Skill.md
SKILL.MD

✓ Correct:

SKILL.md

Vague Descriptions

❌ Wrong:

description: Helps with testing

✓ Correct:

description: Run Playwright-based web application tests. Use when validating UI functionality or running end-to-end test suites.

Missing File References

❌ Wrong: Reference a script file that doesn't exist in the skill directory.

✓ Correct: Only reference files that are actually bundled with the skill.

Best Practices

1. Test Early: Validate during development, not just before deployment
2. Fix Errors First: Address all failures before warnings
3. Iterate: Test after each change
4. Read Reports: Don't just check pass/fail, read the details
5. Automate: Integrate into CI/CD pipelines

Validation Checklist

Use this before deploying skills:

• Run python scripts/test_skill.py /path/to/skill
• Zero errors in validation report
• Warnings addressed or documented
• All file references valid
• Description clear and actionable
• Skill name matches directory
• YAML frontmatter valid

Resources

• Validation script: scripts/test_skill.py
• YAML validator: scripts/validate_yaml.py
• Agent Skills Specification: https://github.com/anthropics/skills/blob/main/agent_skills_spec.md