User: "Create a skill for validating JSON schemas"

You: [Use this skill to generate SKILL.md]

.claude/skills/my-skill/
├── SKILL.md              # Required, uppercase
├── examples/             # Optional
│   └── usage.md
├── templates/            # Optional
│   └── template.yaml
└── scripts/              # Optional
    └── helper.py

---
name: skill-name-lowercase-hyphenated
description: What the skill does, when to use it, trigger keywords. Max 1024 characters.
allowed-tools: Read, Grep, Glob  # Optional: restrict tool usage
---

description: Validate JSON documents against JSON Schema specifications. Use when working with JSON validation, API request/response validation, or data integrity checks. Supports draft-07, draft-2019-09 schemas. Triggers on "JSON schema", "validate JSON", "schema validation".

description: Helps with Excel files  # Too vague

---
name: extracted-name
description: [Comprehensive description]
---

# Skill Name

## Purpose
Brief explanation

## When to Use
- Use case 1
- Use case 2

## Core Instructions
Step-by-step guidance

### Patterns and Examples
Code examples

## Dependencies (Optional)
Required tools/libraries

## Version
v1.0.0

For examples, see [examples/usage.md](examples/usage.md)

---
name: safe-file-reader
description: Read files without making changes
allowed-tools: Read, Grep, Glob
---

⚠️ This skill enforces discipline. Run skill-hardening to test it:
- Create pressure scenarios
- Verify agents comply under stress
- Close loopholes

---
name: json-schema-validator
description: Validate JSON documents against JSON Schema specifications. Use when working with JSON validation, API request/response validation, configuration file validation, or data integrity checks. Supports draft-07, draft-2019-09, and draft-2020-12 schemas. Triggers on "JSON schema", "validate JSON", "schema validation", "JSON structure check".
---

# JSON Schema Validator

## Purpose
Validate JSON documents against JSON Schema specifications.

## When to Use
- Validating API payloads
- Checking configuration files
- Data integrity verification

## Core Instructions

1. **Identify Schema**: Locate the JSON Schema
2. **Load JSON**: Read the document to validate
3. **Validate**: Compare against schema rules
4. **Report**: List validation errors with paths

### Basic Validation Pattern

```python
import jsonschema
import json

def validate_json(data, schema):
    try:
        jsonschema.validate(instance=data, schema=schema)
        return True, "Valid"
    except jsonschema.exceptions.ValidationError as e:
        return False, str(e)

## Example: Multi-File Skill

**Request:** "Create a skill for API testing with examples"

**Generated structure:**

## Best Practices

### Keep It Minimal
- Don't create files just because you can
- Each supporting file should add value
- Examples should teach, not overwhelm

### Description Writing
- Start with action verbs: "Analyze", "Generate", "Transform"
- Include specific file types: ".xlsx", ".json", ".py"
- List concrete use cases
- End with explicit triggers

### Progressive Disclosure
Supporting files load on-demand. Reference them:
```markdown
See [examples/usage.md](examples/usage.md) for detailed examples.