Fire

Incident management for teams that want clear, durable incident state with low operational overhead.

Architecture

Slack / Dashboard actions
          |
          v
     services/incidentd
 (Cloudflare Worker + Durable Objects + Workflows)
          |
          +--> services/dashboard   (operator UI)
          +--> services/status-page (public HTML status pages)
          +--> packages/db          (shared Postgres schema)

Services

services/incidentd

Core incident runtime.

• Durable Object is source of truth for incident state/events.
• Outbox + alarms forward events to workflows.
• Workflows dispatch side effects (D1 index updates, Slack/status-page updates, etc.).

services/dashboard

SolidJS operator dashboard.

• Incident and configuration UI.
• Server functions with auth middleware.
• TanStack Router + Query patterns documented in services/dashboard/AGENTS.md.

services/status-page

Next.js service that returns public status page HTML.

• Domain-based page resolution.
• Incident history pages and RSS/Atom feeds.
• Reads status-page data from shared DB schema.

packages/db

Shared Drizzle schema package (@fire/db) used across services.

Getting Started

bun install
bun run dev

# Run services individually
bun run dev:dashboard
bun run dev:incidentd
bun run dev:status-page

Environment

Each service uses its own .env file. See service READMEs for required variables.

License

MIT