feat(03-01): create MetricsCollector and wire into gateway

- Add MetricsCollector class with counters, model call ring buffer, event ring buffer, and active request tracking
- Add system.metrics, system.events, system.activeRequests RPC handlers
- Add GET /health unauthenticated HTTP endpoint for Docker HEALTHCHECK
- Add totalPending() to LaneQueue for queue depth metrics
- Add 20 tests for MetricsCollector

src/gateway/handlers/system.ts

@@ -1,5 +1,6 @@
 import type { GatewayRequest, OutboundMessage } from '../protocol.js';
 import { makeResponse, makeError, ErrorCode } from '../protocol.js';
+import type { MetricsSnapshot, EventEntry, ActiveRequestInfo } from '../metrics.js';
 
 /** Per-session token usage report returned by system.tokenUsage. */
 export interface TokenUsageEntry {
@@ -21,6 +22,12 @@ export interface SystemHandlerDeps {
   getUsage?: () => { totalSessions: number; activeConnections: number };
   /** Optional callback to retrieve per-session token usage data. */
   getTokenUsage?: () => TokenUsageEntry[];
+  /** Optional callback to retrieve aggregated metrics snapshot. */
+  getMetrics?: () => MetricsSnapshot;
+  /** Optional callback to retrieve recent events. */
+  getEvents?: (opts?: { level?: string; limit?: number }) => EventEntry[];
+  /** Optional callback to retrieve active requests. */
+  getActiveRequests?: () => ActiveRequestInfo[];
 }
 
 export function createSystemHandlers(deps: SystemHandlerDeps) {
@@ -75,5 +82,28 @@ export function createSystemHandlers(deps: SystemHandlerDeps) {
       const sessions = deps.getTokenUsage?.() ?? [];
       return makeResponse(request.id, { sessions });
     },
+
+    'system.metrics': async (request: GatewayRequest): Promise<OutboundMessage> => {
+      if (!deps.getMetrics) {
+        return makeResponse(request.id, {});
+      }
+      return makeResponse(request.id, deps.getMetrics());
+    },
+
+    'system.events': async (request: GatewayRequest): Promise<OutboundMessage> => {
+      if (!deps.getEvents) {
+        return makeResponse(request.id, { events: [] });
+      }
+      const params = request.params as { level?: string; limit?: number } | undefined;
+      const events = deps.getEvents({ level: params?.level, limit: params?.limit });
+      return makeResponse(request.id, { events });
+    },
+
+    'system.activeRequests': async (request: GatewayRequest): Promise<OutboundMessage> => {
+      if (!deps.getActiveRequests) {
+        return makeResponse(request.id, { requests: [] });
+      }
+      return makeResponse(request.id, { requests: deps.getActiveRequests() });
+    },
   };
 }