Changelog

April 2026

A public REST API and the first monoscope CLI release, a Sessions view, clearer Slack alerts, and quieter issues.

  • CLI, MCP & agents
  • Logs & traces
  • Monitors & alerts
  • Issues
  • Integrations
  • Onboarding & billing
  • Self-hosting

Highlights

  • A public REST API and the monoscope CLI: search events, query metrics and manage monitors, dashboards and issues from your terminal, with skills for Claude Code.
  • A Sessions view groups spans by user session, with users, errors and duration at a glance.
  • Slack and Discord alerts are shorter, show the top stack frame, and say how long a problem has been firing.
  • Browser errors join backend errors, and framework errors roll up into one issue instead of one per route.
The monoscope explorer with logs, traces and the query editor.

CLI, MCP & Agents

  • REST API. /api/v1 with API-key auth covers monitors, dashboards, API keys, events, metrics, issues, log patterns, endpoints, projects, teams, members and share links.
  • CLI. The first monoscope CLI release: commands for each of those resources, YAML definitions you can apply, shell completion, prebuilt binaries and an install script.
  • Skills for coding agents. A Claude Code skills plugin and full CLI docs.

Logs & Traces

  • Sessions view. Spans grouped by session, with user, event and error counts, duration and a sparkline. Expand a session into its traces. A summary strip shows sessions with errors and p95 duration, with a chart of clean and failed sessions over time.
  • Identity across a trace. A user or tenant set on one span carries to every span in the trace. Span details show name, email, user ID and session ID.
  • Query editor. Suggestions from your real values, a library of saved and popular searches, and inline parse errors.
  • Waterfall. Spans are labeled with the query, route or RPC method, and the timeline resizes. Browser spans are labeled as page load, fetch, navigation, click, long task or web vital.
  • Spans whose parent never arrived group under one placeholder, and latency bars are corrected for clock skew.

Monitors & Alerts

  • Clearer alerts. Slack and Discord alerts carry a severity color, the top stack frame, a View trace button and signed percentage changes.
  • Still firing. Repeat alerts say how long the problem has lasted, without duplicates.
  • Teams are alerted for new log pattern, rate change and runtime error issues, not only in the inbox. Remove one person from alert emails without removing them from the project.
  • Monitors use their configured time window, recover when results are empty, and show the right current value.

Issues

  • Browser errors. Errors from browser SDKs go through the same issue pipeline as backend errors.
  • Rolled-up errors. Connection resets and similar framework errors become one issue instead of one per route.
  • Quieter inbox. Groups of similar log pattern issues collapse into one row, and stale discovery issues archive after 14 days.
  • New endpoint pages. Request trend, endpoint details, field changes and the originating trace.
  • The API catalog has Incoming, Outgoing and Archived tabs, accurate per-host counts, and the services that serve or call each host.
  • User clicks, console events and navigation before an error show in the issue’s activity panel.

Session Replay

  • The play button has a red border when the session had errors. Missing pieces no longer break a whole recording.

Integrations

  • Slack alerts can go to private channels, and chart images render inline.

Onboarding & Billing

  • New paid subscriptions start with a 30-day trial, with reminders 7 and 3 days before it ends.
  • The sidebar has flyout menus for Explorer, API Catalog, Issues, Monitors and Settings.
  • Deleting a project asks for confirmation.

Self-Hosting

  • DATABASE_URL accepts keyword/value connection strings such as host=... sslmode=disable.
  • Crashed service threads restart on their own.

Know what happened. Prove it.