Skip to main content

MCP Use Cases

This page provides practical examples of how to interact with WIXY using an AI agent.

Available Tools Reference​

Your AI agent automatically understands these tools through their JSON Schema:

  • listStubs: Get all active mappings.
  • createStub(json): Add a new mapping.
  • deleteStub(id): Remove a mapping.
  • enableProxy(url): Turn on upstream forwarding.
  • disableProxy: Turn off forwarding.
  • startRecording(url): Capture traffic.
  • stopRecording: Save captured traffic as stubs.

👨‍💻 For Developers: Rapid Scaffolding​

Scenario: You are building a checkout page and the payment service isn't ready.

  • User: "Claude, I need a mock for the payment service. It's a POST to /api/v1/pay. It should accept a JSON with amount and currency. If successful, return a 201 Created with a transaction ID like TXN-XXXX."
  • AI Action: Generates WireMock JSON and calls createStub.
  • AI Result: "I've created the payment mock for you. You can now send POST requests to port 9090."

🧪 For QA: Environment Management​

Scenario: You need to capture a complex state from a staging environment.

  • User: "Start recording traffic from https://staging.acme.com. I'm going to run the 'Forgot Password' flow now."
  • User Action: (Performs the UI steps)
  • User: "Stop recording and tell me what you captured."
  • AI Action: Calls stopRecording, audits the new stubs.
  • AI Result: "Recording stopped. I captured 4 stubs including /api/auth/reset and /api/mail/sent. They are now active in WIXY for offline use."

🛡️ For DevOps: Resilience Testing​

Scenario: You want to see how the system handles a partial outage.

  • User: "I want to test our 'circuit breaker' implementation. Make the inventory service at /api/inventory/.* return a 503 error with a 2-second delay for every request."
  • AI Action: Creates a high-priority stub with fixedDelayMilliseconds and a 503 status.
  • AI Result: "The inventory service is now simulating a slow failure state. You can verify your circuit breakers now."

🤖 Advanced: Self-Healing Tests​

AI agents can use WIXY's tools to fix test environments autonomously:

  1. AI: "I see the integration test for 'User Profile' failed with a 404."
  2. AI: (Calls listStubs) "The stubs are present, but the path is /v1/users while the code is calling /v2/users."
  3. AI: (Calls createStub) "I've added a mapping for /v2/users to match the new API version. Retrying tests..."