A running log of new features, improvements, and fixes across the GetFluxly platform: Console, SDK, and HTTP API.
This release ships multi-surface project tracking, live/test project modes, a custom analytics time-range picker, smarter onboarding defaults, and official client libraries for six platforms.
Every item below is live in the Console and the HTTP API.
GetFluxly now has paid plans. Subscribe via Dodo Payments, start a 14-day trial that grants Growth-level access with no credit card required, and manage everything from the new Plan and billing page in account settings. Live usage meters show your consumption against each plan limit as you go.
A project can now contain multiple surfaces. Each surface gets its own SDK API keys, its own scoped analytics view, and its own default event properties that the ingest pipeline merges automatically. A contextual switcher in the Console header lets you move between surfaces without leaving the page.
A focused security pass covering several surface areas identified during a pre-launch audit.
Projects now carry an explicit mode: live or test. Switching modes requires confirming a modal dialog so accidental flips are prevented. Test-mode events are stored separately and do not affect live analytics or automation triggers.
Break any funnel down by acquisition source and see revenue attributed by channel. Last-touch attribution is the default. Channel classification reuses the same classify_channel logic applied at ingest so the breakdown is consistent across every analytics view.
Every analytics dashboard now accepts a custom date range in addition to the built-in presets. Pick any start and end date to scope funnels, retention, and event counts to the window that matters for your analysis.
Build multi-step conversion funnels and track retention cohorts, now grouped together under a dedicated Product Analytics section. Define each funnel step by event name and optional property filters, then view drop-off at every stage. Retention cohorts show return rates by day and week for any anchor event.
New projects now seed a default surface during onboarding and bind the first SDK key to it automatically. You reach the Console with a working key already in place rather than needing to create the surface and key manually before your first event arrives.
The email editor now renders the auto-appended compliance footer inline so you see exactly what recipients will see before you send. Sender details, such as your company address and unsubscribe link text, can be configured without leaving the editor.
New accounts now go through a guided post-auth onboarding flow that captures intent and configures the project around it. An intent-aware setup hub in the Console surfaces the right next steps for your use case, and a sidebar setup-progress rail tracks completion of your first-run checklist.
GetFluxly now uses email-first magic-link sign-in by default. Enter your email, click the link, and you are in. The whole flow completes in the same browser tab with no password required. A one-time code option is available on request for environments where links are blocked.
GetFluxly now ships official client libraries for six platforms. Every library wraps the same HTTP API, handles authentication via the x-gflux-key header, and exposes a consistent track/identify/page surface.
@getfluxly/browser: auto-captures page views, anonymous identity, and consent; exposes window.gflux.@getfluxly/react: provider + hooks wrapping the browser SDK with first-class React lifecycle integration.@getfluxly/next: SSR-safe provider that handles client/server event routing and App Router compatibility.@getfluxly/node: server-side batch ingest with automatic retry and backoff.getfluxly on PyPI: sync and async clients for Django, FastAPI, and script use.getfluxly on RubyGems: thread-safe background flush with Rack middleware helpers.Dashboards and timestamps across the Console now render in your workspace timezone instead of raw UTC. Set your preferred timezone once in workspace settings and every chart label, event timestamp, and scheduled time reflects it.
You can now permanently delete an individual project or your entire account from account settings. Before the operation runs, the Console shows a clear summary of which workspaces will be affected. A full cascade cleanup removes all underlying data once confirmed.
An outbound webhook delivery pipeline lets you forward GetFluxly events to any HTTPS endpoint. Configure retry behavior, inspect delivery logs, and toggle endpoints on or off per event type. A companion Stripe inbound source connects your Stripe account and runs a one-shot historical backfill of customers, subscriptions, and charges so your profile data is complete from day one.
Historical analytics queries now hit pre-computed daily rollups instead of scanning raw event rows, so dashboards for longer time windows load significantly faster. Write-side hygiene improvements apply bot detection, spam filtering, and consent-drop counters at ingest time, keeping rollup data clean without a backfill step.
The live visitor map on the overview now renders as a choropleth: regions shade darker as active visitor count rises. At a glance you can see where traffic is concentrated right now without reading individual pin counts.
After eight months of building in the dark, the GetFluxly private beta opens on Monday, May 25, 2026. Accepted teams get three months of full platform access at no cost, with no card on file. One platform for product events, unified profiles, behavioral segmentation, lifecycle automation, and email send, behind one SDK and one HTTP API.
Full details, FAQ, and the application link are in the launch announcement.
The browser SDK's 0.4.0 release is a focused hardening pass on the pieces that send events and store identity. The headline fix is in the transport: the old delivery path treated any HTTP response as success so long as the network call resolved, which meant 4xx and 5xx responses were silently dropped and never retried. The transport now inspects response.ok and response.status, retries 408/425/429/5xx with jittered exponential backoff, and surfaces non-retryable rejections through a pluggable logger config. No code changes required; affected workspaces will see ingest completeness climb on the next deploy.
sendBeacon only. Regular interval flushes used to share the 64KB per-origin keepalive budget and threw under real volume. Standard flushes now use a regular fetch; keepalive is reserved for page-leave delivery via sendBeacon, where it actually matters./v1/sdk-config responses are no longer trusted blindly. Unknown fields are dropped, invalid enums are rejected, and sampling is clamped to [0, 100]. A compromised or misconfigured config endpoint cannot flip the SDK's privacy defaults out from under the customer.Secure-flagged on HTTPS. The cookie fallback used by the identity layer now sets the Secure attribute whenever the page is served over HTTPS, so it cannot leak over plain HTTP.apiHost validated with new URL(). Non-http(s) protocols are rejected; URLs that already include a path are preserved as-is rather than having endpoint segments concatenated naively. Misconfiguration fails loudly at init instead of corrupting URLs in production.consentBanner.nonce or set <meta name="gflux-csp-nonce"> and the injected <style> tag runs under your CSP without needing style-src: unsafe-inline.pageview and page-leave previously detected route changes independently and raced; both now subscribe to a single shared SPA router notifier. page-leave fires on SPA navigation in addition to visibility and unload.page_view. Not just session_start.gflux.destroy() is now async. Returns a Promise<void>. Fire-and-forget callers still work; callers that need delivery guarantees on teardown can await it to wait for the transport to drain.You can now select any profile in the Console, pick an automation, and run it against that user's full event history as if today were a different date. Useful for debugging "why didn't this fire?" and for testing new flows against real cohorts before they go live.
The segment builder used to recompute counts on save. It now streams an estimated count as you add each filter, with the full count finalising in < 2s for most workspaces. Caching is per-workspace, scoped to the current event window.
If your app called window.gflux.identify() in the same tick as the SDK boot, the identify could land before the SDK had read its anonymous ID from storage, producing an orphan profile. The SDK now buffers identify calls until init has fully resolved.
The Events page now has a tail-mode that streams new events as they arrive over a long-lived SSE connection. Default page-load still shows the last 24 hours; press L to switch to live tail.
L tail, P pause, Esc exit.The HTTP Events API docs now call out that ingest is at-least-once. Server callers should retry 5xx responses with backoff and tolerate duplicate events after unknown network outcomes.
/v0 endpoints retired.The /v0 namespace, which has been deprecated since November, is now removed. All endpoints live under /v1. Beta workspaces have already been migrated automatically; if you have client code calling /v0/track, swap the path to /v1/track. The request and response shapes are identical.