Support multiple teams per alert

OPERATIONS.md

@@ -14,8 +14,8 @@ We map extra fields using what Grafana calls "Custom annotation name and content
 These must be present on every alert for it to fire:
 
 They're basically key/value pairs:
-- **`team`** - The owning team identifier (e.g., `pytorch-dev-infra`, `pytorch-benchmarking`)
-- **`priority`** - Alert severity level: `P0`, `P1`, `P2`, or `P3`
+- **`Teams`** - The owning team identifier(s). Supports multiple teams separated by commas (e.g., `pytorch-dev-infra, pytorch-benchmarking`)
+- **`Priority`** - Alert severity level: `P0`, `P1`, `P2`, or `P3`
 
 ### Optional Annotations
 
@@ -30,20 +30,20 @@ These fileds will also be populated in the alerts
 ### Configuration
 
 1. Create your alert rule in Grafana.  Give the outputs of the query meaningful names, otherwise Grafana will default to A, B, C
-2. Add the required fields in labels:
+2. Add the required fields in annotations:
    ```
-   team = dev-infra
-   priority = P1
+   Teams = dev-infra, platform
+   Priority = P1
    runbook_url = https://wiki.example.com/runbooks/disk-space
    ```
 3. Under "Configure Notification" enable "advanced options". Alerts will now get routed to our Dev and Prod channels
 
 ### Example Configuration
 
 ```yaml
-labels:
-  team: "dev-infra"
-  priority: "P1"
+annotations:
+  Teams: "dev-infra, platform"
+  Priority: "P1"
   runbook_url: "https://wiki.pytorch.org/runbooks/disk-space"
 ```
 
@@ -55,7 +55,7 @@ CloudWatch alerts use the AlarmDescription field to pass metadata in a specific
 
 These must be present in the AlarmDescription:
 
-- **`TEAM`** - Owning team identifier
+- **`TEAMS`** - Owning team identifier(s). Supports multiple teams separated by commas
 - **`PRIORITY`** - Priority level (`P0`, `P1`, `P2`, `P3`)
 
 ### Optional Fields
@@ -68,7 +68,7 @@ The AlarmDescription should contain your alert description, followed by metadata
 
 ```
 High CPU usage detected on production instances
-TEAM=dev-infra
+TEAMS=dev-infra, platform
 PRIORITY=P1
 RUNBOOK=https://wiki.pytorch.org/runbooks/high-cpu
 ```
@@ -91,7 +91,7 @@ When properly configured alerts fire:
 
 The resulting GitHub issue includes:
 - Normalized title and description
-- Team and priority labels
+- Multiple team labels (Team:dev-infra, Team:platform, etc.) and priority labels
 - Links to runbooks if provided
 - Debug information from original alert payload
 

README.md

@@ -386,7 +386,7 @@ interface AlertEvent {
   reason?: string;              // Provider-specific reason/message
   priority: "P0" | "P1" | "P2" | "P3"; // Canonical priority
   occurred_at: string;          // ISO8601 timestamp of state change
-  team: string;                 // Owning team identifier
+  teams: string[];              // Owning team identifiers (supports multiple teams)
   resource: {                   // Resource information
     type: "runner" | "instance" | "job" | "service" | "generic";
     id?: string;                // Resource identifier
@@ -421,7 +421,7 @@ interface AlertEvent {
   "summary": "Critical CPU alert on production web server",
   "priority": "P1",
   "occurred_at": "2024-01-15T10:30:00Z",
-  "team": "platform-team",
+  "teams": ["platform-team"],
   "resource": {
     "type": "instance",
     "id": "i-1234567890abcdef0",
@@ -512,20 +512,25 @@ aws secretsmanager update-secret \
 
 **CloudWatch Alarms** - Add to AlarmDescription:
 ```
-TEAM=dev-infra | PRIORITY=P1 | RUNBOOK=https://runbook.example.com
 High CPU usage detected on production instances.
+TEAMS=pytorch-dev-infra, pytorch-platform
+PRIORITY=P1
+RUNBOOK=https://runbook.example.com
 ```
 
-**Grafana Alerts** - Use labels:
+**Grafana Alerts** - Use annotations:
 ```yaml
-labels:
-  team: dev-infra
-  priority: P2
 annotations:
+  Teams: pytorch-dev-infra, pytorch-platform
+  Priority: P2
   runbook_url: https://runbook.example.com
   description: Database connection pool exhausted
 ```
 
+**Multi-Team Support**: Use comma-separated teams:
+- Single team: `TEAMS=dev-infra` or `Teams: dev-infra`
+- Multiple teams: `TEAMS=dev-infra, platform, security` or `Teams: dev-infra, platform, security`
+
 ## 📋 Operations Guide
 
 For detailed instructions on configuring new alerts in Grafana and CloudWatch, see [OPERATIONS.md](OPERATIONS.md).
@@ -610,7 +615,7 @@ All logs use structured JSON with correlation IDs:
   "messageId": "12345-abcde",
   "fingerprint": "abc123...",
   "action": "CREATE",
-  "team": "dev-infra",
+  "teams": ["pytorch-dev-infra", "pytorch-platform"],
   "priority": "P1",
   "source": "grafana"
 }
@@ -627,14 +632,19 @@ All logs use structured JSON with correlation IDs:
 4. Look for circuit breaker or rate limiting logs
 
 **Missing required fields error:**
+
+CloudWatch alerts need TEAMS and PRIORITY in AlarmDescription
 ```bash
-# CloudWatch alerts need TEAM and PRIORITY in AlarmDescription
-TEAM=dev-infra | PRIORITY=P1 | RUNBOOK=https://...
+TEAMS=dev-infra, platform
+PRIORITY=P1
+RUNBOOK=https://...
+```
 
-# Grafana alerts need team and priority labels
-labels:
-  team: dev-infra
-  priority: P2
+Grafana alerts need Teams and Priority annotations
+```
+annotations:
+  Teams: dev-infra, platform
+  Priority: P2
 ```
 
 **High DLQ depth:**

lambdas/collector/__tests__/basic-functionality.test.ts

@@ -31,7 +31,7 @@ describe("Basic Functionality Tests", () => {
       expect(result.source).toBe("grafana");
       expect(result.state).toBe("FIRING");
       expect(result.title).toBe("Test Alert");
-      expect(result.team).toBe("dev-infra");
+      expect(result.teams).toEqual(["dev-infra"]);
       expect(result.priority).toBe("P1");
     });
 
@@ -63,7 +63,7 @@ describe("Basic Functionality Tests", () => {
       expect(result.source).toBe("cloudwatch");
       expect(result.state).toBe("FIRING");
       expect(result.title).toBe("High CPU Usage");
-      expect(result.team).toBe("platform");
+      expect(result.teams).toEqual(["platform"]);
       expect(result.priority).toBe("P2");
     });