v2.2.0 - Structured Logging + Prometheus Metrics

AgentWire CORE v2.2.0

Observability infrastructure for Phase 0 stability.

New Features

Prometheus Metrics Endpoint

• GET /metrics — Prometheus scrape format (no auth required)
• GET / — root endpoint now lists metrics path

Metrics Exported

Metric	Type	Labels	Purpose
agentwire_router_hit_total	Counter	rule_name	Routing rule hit count
agentwire_routing_duration_seconds	Histogram	—	Routing decision latency
a2a_dispatch_total	Counter	peer, status	Dispatch attempts (success/error)
a2a_message_duration_seconds	Histogram	peer, agent_id, rule_name	End-to-end message latency
a2a_task_state_total	Counter	state	Task state transitions

Structured Logging

• Added structlog dependency (>=24.1.0)
• _configure_structlog() helper for JSON-formatted logs

Dependencies

• +structlog>=24.1.0
• +prometheus-client>=0.20.0

Testing

• 31 tests passing (added 5 new test classes for metrics instrumentation)
• TestAgentCard, TestAgentRouter, TestSqliteTaskStore, TestGatewayExecutor, TestBearerTokenMiddleware
• TestMetricsEndpoint, TestRouterMetrics, TestExecutorMetrics, TestTaskStoreMetrics (new)

Files Changed

• server_v2/__init__.py — version bump 2.1.0 → 2.2.0
• server_v2/requirements.txt — +structlog, +prometheus-client
• server_v2/agent_router.py — metrics instrumentation
• server_v2/gateway_executor.py — metrics instrumentation
• server_v2/task_store.py — metrics instrumentation
• server_v2/start.py — /metrics endpoint + root endpoint update
• tests/test_v2_core.py — +5 test classes

Alignment with Strategic Goals

This release supports Phase 0 (~2029) stability requirements by providing:

• Production-grade observability via Prometheus metrics
• Foundation for Grafana dashboards and alerting
• Structured logs for log aggregation pipelines

---

Commit: 273ccbb
Author: silk-thread silk-thread@agentwire.local