Breakpoints — pause and inspect, on either side

A Breakpoint catches a live request — or its response — on a matching host before it's delivered, so you can inspect it in full before deciding what happens next. On a request, you can edit the method, path and query parameters, headers, or body. On a response, you can edit the status, headers, or body. Once you're done, choose one of three actions: Forward it, edited or unmodified; Reset to Original to discard your edits and go back to exactly what was captured; or Abort to end the exchange outright. Use this when you want to intervene in a specific moment rather than set up something that runs automatically every time.

Rules — the same intervention, automated

A Rule does what a Breakpoint does, but repeatably and without you tapping through it live: redirect, block, throttle, mock a response, or rewrite headers, matched by host and path. Matching supports both wildcards and full regular expressions, not just a plain substring match, so a rule can target exactly api.example.com/v1/* without also catching v10.

Rules apply across HTTP/1.1, HTTP/2, and tunneled connections uniformly — a rule you write once behaves the same regardless of which protocol a given app happens to use.

Mocking — replace the response entirely

Mocking is a Rule that answers with a response you control — a specific status, headers, and body — instead of forwarding to the real server. Use it to test how your app behaves against a server state you can't otherwise produce on demand, like a 500 from a backend that's actually healthy, or a payload shape the real API doesn't return yet. The fastest way to build one: open a captured response's detail view and tap Create Mock Rule, which starts from exactly what was captured, so you're adjusting a real response rather than writing one from scratch.

Network Chaos — simulate a bad network in one tap

Under More → Network Chaos, one-tap presets bundle latency, packet loss, and server errors: Poor Network, Offline API, Slow Backend, Server Errors, and Auth Failure. Each is the equivalent of hand-configuring several Rules at once, for the common case of "how does my app actually behave when the network is bad," without building that configuration yourself every time.

Which one to use

  • Testing one specific request, once — Breakpoint.
  • The same intervention every time a request matches — Rule.
  • Testing against a specific response you control, repeatably — Mocking.
  • General resilience testing, not one endpoint in particular — Network Chaos.