Developer tools

Developer tools is the integration cockpit: keys, webhooks, logs, and simulators. Everything here scopes to the test/live toggle, so you always know which environment you are inspecting.

API keys

The Keys screen manages both key types across both modes — publishable (pk_test_…/pk_live_…) for client-side SDKs and secret (sk_test_…/sk_live_…) for your servers. See API keys for usage and security practices.

• Create — keys are named ("Production backend", "CI"), and the secret is shown once at creation. Copy it then; afterwards only a masked prefix is visible.
• Roll — generates a replacement while the old key keeps working through a grace period you choose, so you can migrate without downtime.
• Revoke — kills a key immediately, with confirmation.
• Last used — every key shows when it last authenticated a request, so stale keys are easy to spot and safe to remove.

Each key row includes copy-to-clipboard and an inline SDK snippet with the key already injected.

Webhook endpoints

The Webhooks screen manages your outbound webhook endpoints — how reward grants and lifecycle events reach your backend.

• Create an endpoint — set the URL and subscribe it to event types from the catalog (referral.validated, reward.pending, payout.available, and the rest). Endpoints are registered per mode.
• Signing secret — each endpoint gets an HMAC secret for verifying the Invitebase-Signature header, shown once at creation and rollable at any time.
• Test delivery — fire any event type at the endpoint from the dashboard with a realistic payload and see the response inline.
• Delivery log — every delivery attempt, with the full request payload, response status and body, timing, and retry history.
• Redeliver — one click resends any delivery, whether it failed or you fixed a bug and want it replayed.

In test mode the retry schedule is compressed (a few retries over minutes rather than the ~3-day live backoff), so failure handling is testable in one sitting.

Webhook scenario simulator

Instead of testing one event at a time, the scenario simulator fires an entire referral lifecycle — linked → in_progress → validated → reward — as a real sequence of webhook deliveries to your endpoint, built from your own test-mode campaigns and participants. Preset scenarios cover the happy path, rejection, and expiry. Deliveries land in the normal delivery log, and the run summary shows which step your handler failed. Test mode only.

Event log

The Events screen is the answer to "did my event arrive and what did it look like?" — a live-updating stream of every ingested event, filterable by event name, user, source (SDK, API, or an integration), and time.

Clicking an event shows the full payload: properties, the idempotency key, which API key received it, and — most usefully — which qualification gates and referrals the event affected. Event names seen here also populate the dropdowns in the gate builder.

API request log

Alongside events, the request log records every API call made with your keys: method, path, status code, timing, request and response bodies, and the key used. Filter by status to find failing calls, or by path to audit a specific integration. Errors link to the relevant error reference.

Test-mode simulators

Two tools let you prove the whole referral loop with no app and no real users.

Event simulator

Fire synthetic events from the dashboard against test-mode users: pick or create a test participant, choose an event — including simulated integration events like a RevenueCat purchase — set properties, and fire. Simulated events flow through the real ingestion path and gate engine; there are no shortcuts. Watch a test referral move linked → in_progress → validated → reward as events land. The simulator is available from the Get started checklist and from Developer tools.

Rule inspector

A step-by-step debugger for qualification gates, evaluated against any simulated or real test user:

• Per-gate breakdown: which events matched, property comparisons with actual vs required values, threshold progress, and window state
• An overall verdict with the exact reason a user does or does not qualify
• Live updates alongside the event simulator — fire an event, watch the evaluation change

The inspector is deep-linkable from any referral's detail view, which makes it the fastest answer to "why is this referral stuck in in_progress?"

What's next