Open-source incident communication platform for service status pages. A self-hosted alternative to Atlassian Statuspage and PagerDuty Status Page.
UI is provided by Garden UI — a separate frontend project built on top of the IncidentGarden API.
The open-source status page landscape is split: monitoring tools (Uptime Kuma, Gatus) detect problems but don't help communicate them to users. Static pages (cState, Upptime) display status but lack incident management. Commercial platforms (Statuspage, BetterStack) do both — for $99-1499/month.
IncidentGarden fills the gap: a full incident communication lifecycle in an open-source, self-hosted package.
- Incident lifecycle with audit trail — not just open/close, but
investigating>identified>monitoring>resolved, with every service change tracked - Effective status auto-computed — worst-case across all active events, per service. No manual status juggling
- Subscriber notifications — users subscribe to specific services and get notified via Email, Telegram, or Mattermost. Not just admin alerts — user-facing communication
- Production-ready from day one — Prometheus metrics, pre-built alerts, Kubernetes probes, structured logging, graceful shutdown. No "add monitoring later"
IncidentGarden is an incident communication platform, not an uptime monitor. It handles what happens after an issue is detected: status tracking, timeline updates, and subscriber notifications.
Pair it with your existing monitoring stack — Prometheus, Uptime Kuma, Grafana, or any tool that detects problems. IncidentGarden handles the communication.
Status Page
- Public status page with real-time service statuses
- Service groups with M:N membership
- 5 status levels:
operational,degraded,partial_outage,major_outage,maintenance - Status history and audit log
Incident & Maintenance Management
- Full incident lifecycle:
investigating>identified>monitoring>resolved - Scheduled maintenance:
scheduled>in_progress>completed - Severity levels:
minor,major,critical - Per-incident affected services with granular status overrides
- Add/remove services on the fly during an active incident
- Event templates for consistent communication
- Complete audit trail of every change (who, when, what)
Notifications
- 3 channels: Email (SMTP), Telegram (Bot API), Mattermost (webhooks)
- Per-service subscriptions — users choose what they care about
- Channel verification (email codes, Telegram /start, Mattermost test message)
- Async delivery queue with retry mechanism
- Default email channel auto-created on registration
Access Control
- Three roles:
user(subscribe) >operator(manage incidents) >admin(full control) - JWT authentication with refresh tokens
Operations
- Prometheus metrics on
/metrics(port 9090) - Pre-configured alerts: error rate, latency, DB pool, memory, goroutine leaks
- Health (
/healthz) and readiness (/readyz) probes - Structured JSON logging via slog
- Graceful shutdown
Requirements: Docker and Docker Compose.
git clone https://github.com/bissquit/incident-garden.git
cd incident-garden
cp .env.example .env
make docker-build && make docker-upVerify it's running:
curl http://localhost:8080/healthz # OK
curl http://localhost:8080/api/v1/status # system status JSONInteractive API docs are available at http://localhost:8080/docs, OpenAPI spec at http://localhost:8080/api/openapi.yaml.
For the full visual experience, set up Garden UI — a frontend that connects to the IncidentGarden API.
docker pull ghcr.io/bissquit/incident-garden:latestAvailable for amd64 and arm64.
| Password | Role | |
|---|---|---|
| admin@example.com | admin123 | admin |
| operator@example.com | admin123 | operator |
| user@example.com | user123 | user |
IncidentGarden is designed as an API-first platform. The backend is a standalone product — Garden UI is one client, but you can build your own.
- OpenAPI 3.0 specification — versioned independently from the application
- Spec served at runtime —
GET /api/openapi.yamlalways returns the current contract - Interactive docs — Swagger UI at
/docs - Consistent response format —
{"data": {...}}for success,{"error": {"message": "..."}}for errors
Build on top of it: custom dashboards, CLI tools, ChatOps bots, Grafana panels, CI/CD integrations.
IncidentGarden vs projects in the same niche — incident communication platforms (not monitoring tools):
| IncidentGarden | Cachet v3 | OpenStatus | |
|---|---|---|---|
| Project status | Active | Stalled (1 maintainer) | Active |
| Incident lifecycle | Full (4 states + audit trail) | Basic | Basic |
| RBAC | user / operator / admin | Partial | No |
| Subscriber notifications | Email, Telegram, Mattermost | Email only | |
| Per-service subscriptions | Yes | No | No |
| Event templates | Yes | Yes (Twig) | No |
| Affected services per incident | Multiple, editable on the fly | 1 component | Multiple |
| API contract | OpenAPI 3.0, versioned (v2.12) | Yes | Yes |
| Self-host complexity | Go binary + PostgreSQL | PHP + Laravel + DB | 6 services |
| License | AGPL-3.0 | BSD-3 | AGPL-3.0 |
Uptime Kuma, Gatus, and similar tools are monitoring solutions, not incident communication platforms. They complement IncidentGarden rather than compete with it.
| Part | Tool/Technologie |
|---|---|
| Language | Go 1.25 |
| HTTP Router | chi |
| Database | PostgreSQL 16 |
| Migrations | golang-migrate |
| Logging | slog (stdlib) |
| Metrics | Prometheus |
| CI/CD | GitHub Actions, GoReleaser |
| Containers | Docker (Alpine 3.19, multi-arch) |
- Deployment Guide — environment variables, Kubernetes manifests, Prometheus setup
- OpenAPI Specification — full API contract
- Changelog — release history
# Local development (PostgreSQL in Docker, app with hot-reload)
make docker-postgres
make dev
# Quality checks
make lint
make test
make test-integration
# Database migrations
make migrate-up
make migrate-down
make migrate-create NAME=add_somethingSee make help for all available commands.
Contributions are welcome. Please create issues and pull requests.