Contract testing and flake hardening

Purpose

Catch payload and behavior regressions before integrating with unstable upstream systems.

Where this is used

• Pre-merge local verification for API-consuming clients.
• Manual smoke checks for response contracts and failure handling.
• Reproducing flaky network conditions in a controlled environment.

Setup and prerequisites

• Mockphine desktop app running with target server configured.
• Server fallback set intentionally (`STRICT` is recommended for contract checks).
• Contract-critical endpoints created in `MOCK` mode.

Step-by-step workflow

1. Set fallback mode to `STRICT` so unknown routes fail fast as `404 Not Mocked`.
2. Create/update endpoint bodies to match expected API contract responses.
3. Start server and run your request scripts against local server URL.
4. Add delay/failure simulation on selected endpoints to test retry paths.
5. Use Live View logs to confirm status, duration, and `served-by` source.
6. Adjust endpoint definitions until checks are stable and deterministic.

Expected result and output

• Contract checks run against stable local responses.
• Unexpected routes fail clearly in strict mode.
• Failure and latency scenarios are reproducible across test runs.

Common issues and fixes

• Requests leak to upstream: verify fallback mode and endpoint modes are not passthrough-capable.
• Checks fail with `strict_404`: add missing routes or fix method/path/query settings.
• Delay/failure has no effect: ensure target endpoint is in `MOCK` mode.