Changelog

What we shipped, when we shipped it.

·v2.6

Forms: a form backend that links to session replay

Point any HTML form at one endpoint and get submissions emailed, spam filtered, and stored in an inbox — with a link to the replay of the visit that produced each one.

  • newForm endpoints: POST to one URL from plain HTML, React, fetch, or curl. No backend code, no database.
  • newSpam filtering that weighs a honeypot, submit timing, content heuristics, bot fingerprints, and IP reputation into a single score. Spam is stored so you can audit it, never emailed, and never billed.
  • newEvery submission links to the session replay of the visitor who sent it, when the recorder is installed on the same site.
  • newSubmissions inbox with read, archive, delete, spam filtering, and per-submission detail.
  • newEmail notifications to confirmed recipients only, with Reply-To set to the submitter so replying just works.
  • newFile uploads on paid plans: up to 5 files and 4 MB per submission.
  • newHosted thank-you and error pages, or redirect visitors to your own — validated against the site's domains so the endpoint cannot become an open redirect.
  • newSend a test submission from the dashboard: runs the real path and reports the spam verdict, exactly who could be emailed, and where visitors land. Does not count against your quota.
  • improvedCountry is shown by full name rather than a two-letter code, in notification emails and across the dashboard.
  • fixedOutgoing email was being rejected because the configured sender address was not a verified SES identity. This affected every notification the platform sends — form submissions, recipient confirmations, team invites, usage alerts, uptime and server alerts, and the contact form.
·v2.5

Server CPU & memory monitoring

The Nevision agent now reports per-host CPU, memory, disk, and network telemetry every 10 seconds. Threshold alerts shipped alongside.

  • newServer agent for Linux x86_64 / ARM64 (Ubuntu, Debian, RHEL, Alpine, Amazon Linux).
  • newPer-host metrics: CPU per core, RSS/cache memory, disk usage and IOPS, network bytes in/out, load average, top processes.
  • newTag-based grouping (env, role, region, cluster) across dashboards and alerts.
  • newThreshold alerts on any metric (e.g. memory.used_pct > 90 for 5m).
  • improvedServer view in dashboard with live charts and 30-day history.
·v2.4

Frontend & backend log management

Console.log/warn/error from the browser unified with backend logs from any service. Search, filter, tail in real time.

  • newFrontend log capture (console.log/warn/error/info).
  • newBackend log ingest via POST /v1/logs.
  • newPino, Winston, Bunyan, structlog transports published to npm/pip.
  • newLive tail with composable filters (level, service, sessionId, free text).
  • newAuto-correlation: every backend log line carries the user's sessionId via the x-nevision-session header.
·v2.3

Stripe billing + free plan

Self-serve checkout with Stripe. New free plan (500 sessions/mo, no credit card). Pro at $39, Business at $99 (20% off annually).

  • newStripe-powered self-serve checkout with PaymentElement.
  • newFree plan: 500 sessions/mo, 1 site, 1 uptime monitor, 100 errors/day, 30-day retention.
  • newAnnual billing toggle with ~17% discount on Pro and Business.
  • improvedUsage alerts at 80% and 100% of plan limits via dashboard banner and email.
  • improvedPayment methods manager with default-card setting.
·v2.2

Real User Monitoring (RUM)

Core Web Vitals from real users, broken down per page and device. Daily regression detection.

  • newWeb Vitals capture: LCP, INP, CLS, TTFB, FCP, plus DOMContentLoaded and Page Load.
  • newPer-page breakdowns with p75 percentile (matching Google's CrUX methodology).
  • newDaily regression alerts when any page's p75 metric crosses 'good' thresholds.
  • newCustom metrics via nevision.recordMetric(name, value).
·v2.1

Uptime monitoring + multi-region probes

1-minute interval HTTP checks from US, EU, and Asia regions. Email alerts on confirmed downtime and recovery.

  • newUptime monitors with 1-minute (Pro/Business) or 15-minute (Free) check intervals.
  • newMulti-region probes with 3-of-N consensus to avoid false positives.
  • newEmail alerts on 3 consecutive failures + recovery alert when monitor returns to healthy.
  • newCustom request methods, headers, expected status code, and body string match.
  • new30-day uptime history and response-time charts per monitor.
·v2.0

Error tracking with session linking

JavaScript error capture with stack traces, fingerprint deduplication, and a one-click jump to the session that produced each error.

  • newAutomatic capture of window.onerror and unhandled promise rejections.
  • newFingerprint-based grouping (message + top 3 stack frames).
  • newSource map upload via dashboard or CI API.
  • newStatus workflow: active → resolved → ignored, with auto-reopen on recurrence.
  • newLinked session replay on every error event.