FlowMock sits between your app and your real staging API. By default, every request passes through to staging unchanged. When you need a specific test state - declined payment, empty cart, slow network - you flip a switch and only your session sees the mock.

Everyone else keeps hitting real staging. No collisions. No "who broke the store?" Slack threads.

What you'll need

A staging API your app already talks to
Access to change your app's API base URL (or environment config)
A FlowMock account with at least one project

Step 1: Create a project

In the dashboard, create a team (if you haven't already), then a project. A project maps to one real staging API - think "Checkout staging" or "Mobile app backend."

Inside the project, create an environment (we also call this a proxy endpoint). Each environment gets its own FlowMock URL, like:

text

https://p.flowmock.dev/your-proxy-id

Step 2: Point your app at the proxy

Swap your app's API base URL from staging to the FlowMock proxy URL. Everything else stays the same - same paths, same headers, same auth tokens.

Your app still talks to staging. FlowMock just sits in the middle, watching and optionally overriding.

Step 3: Send a session key

Add a header so FlowMock knows which tester you are:

http

x-flowmock-session: qa-alice

Each person (or CI run) uses their own key. Alice's mocks don't affect Bob's session. Learn more about sessions →

Step 4: Create your first mock

In the dashboard, open your environment and create a mock (we call them overrides). Pick a route - say GET /api/cart - and define the response you want.

Publish it, then activate it for your session. Hit the route from your app. You should see your mocked response instead of staging's.

Step 5: Watch the traffic

Open the Traffic tab. Requests show up immediately for every session on the environment - you do not need to claim a session first.

Session avatars on each row show whose request it was
Use the Sessions filter to narrow the feed, or claim You in the session strip to focus on your own traffic (and so mocks apply to you)

Every row records method, path, status, timing, and whether a mock applied. Perfect for "wait, what did the API actually return?"

What's next?

How it works - the proxy mental model, explained without jargon
QA sessions - claim identity vs watch all traffic
Traffic - live feed, filters, focused mode, and AI edit on captured bodies
Mocks and overrides - matchers, drafts, and the mock library
Transforms - patch live response bodies without full mocks
Scenarios - Record scenario from a click-through, or bundle mocks into presets
Public API - automate session setup from CI