Alert Triage (local)

The alert-triage example is the capstone (L5) of the getting-started golden example. It is a Temporal-driven incident-triage app: an alert comes in, a phased workflow classifies it, gathers context, proposes a remediation, pauses for human approval when the change is risky, applies it (a clearly-stubbed step), then notifies. chant synthesizes the Kubernetes manifests; the triage workflow itself is hand-written Temporal — the “Raw Temporal + chant” split.

This tutorial runs the whole thing locally with no cloud and no cluster — just a Temporal dev server.

Prerequisites

• Node.js and npm.
• The Temporal CLI (temporal) for the dev server.

Run it

cd examples/alert-triage
npm install
npm run dev

npm run dev starts a Temporal dev server, the triage worker, and the webhook receiver, then sends one demo alert. Open the Temporal UI at http://localhost:8233 — you’ll see an alertTriage workflow. The demo alert is high-severity, so it’s risky and pauses at the approval gate.

Approve it:

temporal workflow signal -n default \
  --query "WorkflowType='alertTriage'" --name approve-remediation

The workflow completes, and the worker logs the outcome (… — applied after approval). A low-severity alert skips the gate and applies directly.

Two event sources

Both feed the same triage workflow:

npm run alert            # external alert via the webhook (POST /alert)
npm run drift -- --demo  # a drift event (the WatchOp/lifecycle counterpart)

The webhook is the receiver the WebApp manifest deploys. The drift source runs chant lifecycle plan --json and triages each drifted resource — out-of-band cluster changes go through the same triage as external alerts. --demo injects a sample drift so you can see the pipeline without making a real change.

The agent (optional)

proposeRemediation is a deterministic stub by default — no key, runs offline. Set ANTHROPIC_API_KEY (and npm i @anthropic-ai/sdk) to have it call Claude; override the model with ANTHROPIC_MODEL (default claude-sonnet-4-6). The first run shows chant, not an LLM.

Deploy the manifests

The app’s Kubernetes surface is typed chant in src/:

npm run build      # → k8s.yaml (plain Kubernetes)
npm run lint
kubectl apply -f k8s.yaml   # e.g. to a local k3d cluster

Note

The manifests reference placeholder images (ghcr.io/intentius/alert-*). Swap in your own builds of the worker and webhook to run the app in-cluster; the local npm run dev above runs them from source without images.

How it fits together

Piece	What it is
src/	chant manifests — webhook (WebApp) and worker (WorkerPool)
activities/triage.ts	the triage activities (the agent — stub + Claude opt-in)
activities/workflow.ts	the Temporal workflow (classify → context → propose → gate → apply → notify)
activities/worker.ts	the worker that registers them
app/	the event sources (webhook, drift) and the demo alert

A time-skipping test (activities/workflow.test.ts) covers the gate behaviour in CI. See the example README for the full layout.