Skip to documentation

Guide · Inspect

Available

See what the client actually sent

Traffic inspection connects an observed request to the operation, fixture, condition, template, state effect, and response that handled it. Use that evidence to debug or promote behavior—not screenshots of an assumed request.

In this guide

  • Find request and matching mistakes quickly
  • Inspect response status, body, headers, and timing separately
  • Understand LLM and webhook-specific exchanges
  • Promote useful traffic without copying secrets

Prerequisites

  • A DotMock workspace you can access
  • The API ID in $API_ID and its published runtime URL in $MOCK_URL when the example calls the mock
On this page

Capture requests sent to the mock

Enable inspection for the workspace and exercise the published mock URL. The inspector records traffic that reaches DotMock; it does not intercept unrelated traffic between your application and a third-party URL.

Read the request as structured data

Inspect method, path, path parameters, query pairs, headers, and parsed body separately. Sensitive headers may be redacted. Copy the generated curl command when you need a reproducible request outside the UI.

Connect the winner to the rendered result

The result separates HTTP status, response body, and response headers. Matching and runtime diagnostics explain why a branch won or why an effect was disabled; they are not mixed into the response body.

Use protocol-aware inspection

REST and GraphQL traffic exposes operation and response selection. gRPC inspection decodes metadata, ordered request/response messages, trailers, status, timing, and mock/proxy source. LLM inspection extracts prompts, model, tools, matched fixture, assistant output, and raw HTTP/SSE/WebSocket frames. Webhook delivery history records event, destination, attempts, signature context, response, and retry state.

Promote a useful request carefully

Use captured traffic as a starting point for an endpoint or fixture, then replace one-off IDs, tokens, timestamps, and personal data with models, request helpers, or explicit examples. Add failure branches and test the new definition before publication.