mockzilla-creator（模拟创造者）

Mockzilla Creator Skill

This skill guides you through creating effective, dynamic, and stateful mocks using Mockzilla's MCP tools.

Core Principles

1. Stateless First: Use create_schema_mock for 90% of use cases. It's faster and easier to maintain.
2. Stateful for Workflows: Use create_workflow_scenario only when the response depends on previous actions (e.g., Auth, Multi-step forms).
3. Consistency is Key: Use Interpolation to link fields (e.g., id and message) and Faker for realistic data and only when it's necessary.
4. Deterministic Logic: Keep transitions and effects simple. Complex logic belongs in the schema or the application, not the mock engine.

🛠️ Tool Selection Guide

🎨 JSON Schema & Faker Patterns

The "Premium" Schema Template

Use this pattern to WOW the user with realistic data:

{
  "type": "object",
  "required": ["id", "user", "metadata", "status"],
  "properties": {
    "id": { "type": "string", "format": "uuid" },
    "user": {
      "type": "object",
      "properties": {
        "firstName": { "type": "string", "faker": "person.firstName" },
        "lastName": { "type": "string", "faker": "person.lastName" },
        "email": { "type": "string", "faker": "internet.email" },
        "avatar": { "type": "string", "faker": "image.avatar" }
      }
    },
    "metadata": {
      "type": "object",
      "properties": {
        "createdAt": { "type": "string", "faker": "date.past" },
        "ip": { "type": "string", "faker": "internet.ipv4" }
      }
    },
    "status": { "type": "string", "enum": ["active", "pending", "archived"] }
  }
}

Passing Arguments to Faker

{
  "amount": {
    "type": "string",
    "faker": { "finance.amount": [100, 5000, 2, "$"] }
  }
}

🔗 Interpolation & Field Referencing

CRITICAL: Use interpolation to make your mocks feel "alive" and consistent, but only when it's necessary.

1. Internal Consistency (Inside Schema)

Reference other generated fields using {$.path}.

{
  "orderId": { "type": "string", "format": "uuid" },
  "summary": { "const": "Confirming order {$.orderId} for {$.user.firstName}" }
}

2. Request Data Echo (In Workflows)

Reference input data in your response or effects using {{input.*}}.

• {{input.body.name}}
• {{input.params.id}}
• {{input.query.search}}

3. State & DB Access (In Workflows)

Reference the mini-DB or scenario state.

• {{state.token}}
• {{db.users[0].name}}

🔄 Workflow Recipes

The "Auth Flow" Recipe

1. POST /login:

- Effect: {"type": "state.set", "raw": {"user": "{{input.body}}", "isLoggedIn": true}} - Response: {"status": "ok", "token": "mock-jwt-{{input.body.username}}"}

1. GET /me:

- Condition: {"type": "eq", "field": "state.isLoggedIn", "value": true} - Response: {{state.user}}

The "CRUD" Recipe

1. POST /items:

- Effect: {"type": "db.push", "table": "items", "value": "{{input.body}}"}

1. GET /items:

- Response: {"data": "{{db.items}}"}

⚠️ Troubleshooting

• Reference Errors: If {$.path} appears literally, ensure the referenced field is defined *before* or at the same level in the JSON object (though Mockzilla resolves after generation, order helps readability).
• Condition Mismatch: Use inspect_workflow_state to verify if your values match exactly (type and casing).
• Schema Validation: Ensure type is present for every property in create_schema_mock.
• Default Limits: Mockzilla defaults maxItems to 5. Explicitly set minItems/maxItems in your schema if you need more.

💎 Premium Templates (Copy-Paste)

🛍️ E-commerce Product

{
  "type": "object",
  "required": ["id", "name", "price", "rating", "image"],
  "properties": {
    "id": { "type": "string", "faker": "string.uuid" },
    "name": { "type": "string", "faker": "commerce.productName" },
    "description": { "type": "string", "faker": "commerce.productDescription" },
    "price": { "type": "string", "faker": "commerce.price" },
    "rating": { "type": "number", "faker": { "number.float": { "min": 1, "max": 5, "fractionDigits": 1 } } },
    "image": { "type": "string", "faker": "image.url" },
    "tags": {
      "type": "array",
      "minItems": 2,
      "maxItems": 4,
      "items": { "type": "string", "faker": "commerce.department" }
    }
  }
}

📊 SaaS User Dashboard

{
  "type": "object",
  "properties": {
    "user": {
      "type": "object",
      "properties": {
        "id": { "type": "string", "faker": "string.uuid" },
        "email": { "type": "string", "faker": "internet.email" },
        "plan": { "type": "string", "enum": ["free", "pro", "enterprise"] },
        "status": { "type": "string", "enum": ["active", "suspended"] }
      }
    },
    "stats": {
      "type": "object",
      "properties": {
        "dailyVisits": { "type": "integer", "faker": { "number.int": { "min": 100, "max": 5000 } } },
        "totalRevenue": { "type": "string", "faker": "finance.amount" }
      }
    },
    "recentActivity": {
      "type": "array",
      "minItems": 3,
      "items": {
        "type": "object",
        "properties": {
          "action": { "type": "string", "enum": ["login", "purchase", "update_profile"] },
          "timestamp": { "type": "string", "faker": "date.recent" }
        }
      }
    }
  }
}

📱 Social Feed Post

{
  "type": "object",
  "properties": {
    "id": { "type": "string", "faker": "string.uuid" },
    "author": {
      "type": "object",
      "properties": {
        "username": { "type": "string", "faker": "internet.userName" },
        "avatar": { "type": "string", "faker": "image.avatar" }
      }
    },
    "content": { "type": "string", "faker": "lorem.paragraph" },
    "likes": { "type": "integer", "faker": { "number.int": { "min": 0, "max": 1000 } } },
    "comments": {
      "type": "array",
      "maxItems": 2,
      "items": {
        "type": "object",
        "properties": {
          "user": { "type": "string", "faker": "internet.userName" },
          "text": { "type": "string", "faker": "lorem.sentence" }
        }
      }
    }
  }
}

🚀 Advanced Workflow Recipes

📄 Pagination

Simulate standardized pagination using input.query.

Variables:

• page: {{input.query.page}} (defaults to 1 if missing is handled by app logic, or use fallback in schema)

Transition:

• Path: /api/items
• Response Body:

{
  "page": "{{input.query.page}}",
  "total": 100,
  "data": "{{db.items}}"
}

*Note: Real slicing requires advanced logic, but echoing the page number helps frontend integration.*

🛑 Error Simulation (Chaos Engineering)

Toggle errors without changing code.

1. Set State: POST /admin/toggle-error -> state.set: {forceError: true}
2. Check Condition:

- Trigger: {"type": "eq", "field": "state.forceError", "value": true} - Response: 500 {"error": "Simulated outage"}

🐢 Network Latency

Mockzilla processes requests instantly. To simulate latency:

• Use mock_delay header if supported by client, or
• Rely on client-side throttling (Network tab in DevTools) for UI testing. *Mockzilla itself does not currently support server-side sleep effects.*

🧪 Testing & Verification

1. Schema Validation:

- Use preview_mock with your schema *before* creating the mock. - Check for [ref:key-not-found] in the output to catch bad interpolation paths.

1. State Verification:

- Use inspect_workflow_state to view the *real* values in db.*. - Compare {{db.users[0].id}} with {{state.currentUserId}} to debug mismatches.

1. Performance:

- Keep schemas simple. Deeply nested arrays with maxItems: 100 can slow down generation. - Default maxItems is 5 for a reason—it's enough for UI testing.