Webhook Synthetic Regression Testing for HTTP API Workflows

Catch HTTP API regressions before users do

This n8n workflow exposes a webhook that runs synthetic regression tests against a target HTTP API–driven workflow. It executes test cases, retrieves execution outputs, uses an AI assertion step to compare actual vs. expected results, stores the outcome, and returns a structured pass/fail test report to the requester.

What this workflow does

• Receives a POST request on a webhook endpoint to start a test run (requires at least a targetWorkflowId).
• Validates the request, sets the target workflow details, and generates default synthetic test cases and expected results when none are provided.
• Executes the target workflow by sending an HTTP request with the synthetic test-case payload.
• Fetches results and status via a second HTTP request to retrieve execution outputs, logs, and run status.
• Runs AI-based assertions by calling an AI assertion endpoint to compare actual outputs against expected outputs and detect regressions.
• Stores the test results (and can update the regression baseline), then returns a final report containing per-case pass/fail and an overall HTTP status.

Use cases

• Regression testing for n8n HTTP API workflows after changes to steps, mappings, or upstream integrations.
• SaaS operators needing automated health checks for workflows that are triggered via HTTP.
• Automation engineers validating deterministic behavior with synthetic payloads and AI-assisted comparisons.

Technical details

• Webhook starter endpoint to trigger test runs via POST.
• HTTP Request steps to call endpoints for: execute, results/status, AI assertion, and store baseline/results (replace placeholder httpbin.org URLs with your real endpoints).
• Nodes used: if, code, wait, webhook, sticky note, http request.
• Add required authentication headers if your execution/assertion/storage services need API keys or OAuth.

Perfect for teams implementing continuous confidence in their HTTP API automation—without manually reviewing every regression.