MCP Test MCP

--- story_id: "0018" title: "Prometheus metrics endpoint" created: "2025-10-09" status: "backlog" dependencies: ["0003"] estimated_complexity: "medium" tags: ["observability", "metrics", "phase2"] --- # Story 0018: Prometheus metrics endpoint ## Description Add a Prometheus metrics endpoint to expose connection statistics, tool call counts, error rates, and response time metrics for monitoring and analysis. ## Acceptance Criteria - [ ] Prometheus client library integrated - [ ] Metrics endpoint exposed (HTTP endpoint for metrics) - [ ] Connection count metrics (active, total, failed) - [ ] Tool call count by tool name - [ ] Error count by error type - [ ] Response time histogram (p50, p95, p99) - [ ] Resource and prompt operation metrics - [ ] Metrics reset on server restart (ephemeral) - [ ] Documentation for metrics endpoint - [ ] Example Prometheus configuration ## Technical Notes **Metrics to expose:** ``` mcp_test_connections_total{transport="streaming-http"} 15 mcp_test_connections_active{transport="streaming-http"} 1 mcp_test_connections_failed_total 2 mcp_test_tool_calls_total{tool_name="add"} 42 mcp_test_tool_calls_total{tool_name="multiply"} 15 mcp_test_errors_total{error_type="timeout"} 3 mcp_test_errors_total{error_type="not_connected"} 7 mcp_test_response_time_seconds{quantile="0.5"} 0.087 mcp_test_response_time_seconds{quantile="0.95"} 0.145 mcp_test_response_time_seconds{quantile="0.99"} 0.234 ``` **Implementation approach:** - Use prometheus_client library - Expose metrics on separate HTTP port (default: 9090) - Configure port via MCP_TEST_METRICS_PORT environment variable - Metrics are process-level, not persisted ## AI Directives **IMPORTANT**: As you work through this story, please mark checklist items as complete `[x]` as you finish them. This ensures that if we need to pause and resume work, we have a clear record of progress. Update the `status` field in the frontmatter when moving between stages (in-progress, ready-for-review, done, blocked). Metrics are optional for Phase 2. Focus on useful operational metrics.