Every capability,
itemized.

HTTP probes on configurable intervals, methods, expected status codes, and latency thresholds. Three consecutive failures open an incident; three consecutive successes close it. SSL-expiry, DNS, keyword, and multi-region checks are on the roadmap — not yet shipped.

Receive incidents from Prometheus Alertmanager, Datadog, CloudWatch (via SNS), Grafana, Sentry, or the generic JSON endpoint. Tenant-scoped tokens, Datadog HMAC verification, CloudWatch SSRF hardening, dedup + normalize + auto-resolve after 24h.

Inbound deploy webhooks record every release with SHA, branch, author, and optional commit list. Every incident is ranked against recent deploys by time proximity and confidence. One cURL line from your CI is all the wiring.

Every candidate deploy scored on temporal proximity, locality (does it touch the affected service?), and blast radius (lines changed, tests skipped). Top-N ranked with a one-line rationale per candidate.

Commit analysis pipeline reads the actual diff — hunks, file paths, structured evidence — not just the commit title. Produces a diagnosis explaining which change likely broke what, with direct quotes from the patch.

Pull spans for the failing request from whichever backend you run: Tempo, Jaeger, Honeycomb, New Relic, Elastic APM, AWS X-Ray, or Lightstep. Shown inline with the suspect commit.

One-click revert for confident reversal candidates; AI-drafted patches for smaller fixes. PRs open through our GitHub App with scoped permissions and signed commits. Honors your branch protection — we do not bypass reviewers.

Outbound incident notifications via Slack incoming webhook, routed per channel. Channel-per-incident war rooms, slash commands, and socket-mode two-way chat are on the roadmap— today we send, we don't listen.

Twilio-backed SMS + voice (TwiML announce + DTMF acknowledgement). Web push via VAPID. US/CA/UK at cost; +30% uplift for other destinations. Hard caps, burst limiters, storm-mute on by default.

Rotations, overrides, escalation chains, timezone-aware calendars. iCal export so Google / Outlook show who's holding the pager. Phone-verified contacts with bcrypt-hashed OTPs.

Every resolved incident drafts a postmortem — summary, timeline, customer impact, root cause, contributing factors, action items, and a link to the fix PR. Edit in-app, then publish.

Public and private pages on custom subdomains. Service status aggregation, timestamped incident updates, 30s revalidation. Subscriber email/webhook notifications and scheduled maintenance windows are on the roadmap — not shipped yet.

GitHub + Google OAuth on every plan. SAML + SCIM via WorkOS on Enterprise. Every admin action recorded in a hash-chained, tamper-evident audit log. Verify the chain at any time; export CSV or NDJSON.

AES-256-GCM for observability credentials, webhook signing secrets, and alert ingest tokens. Alert tokens stored as SHA-256 hashes — plaintext returned once, never persisted. Single ENCRYPTION_KEY per deployment today; per-tenant KEKs via KMS are on the roadmap for Enterprise.

Docker Compose for a production-ready install. Bring your own Postgres 15+ and Redis 7+. Helm charts for Kubernetes are on the roadmap — today, you drive self-host through compose.