What problem does it solve? Investigating slow requests and errors across microservices requires digging through distributed traces, and doing it manually through the Jaeger UI is slow and error-prone. This Skill provides scripted, statistics-first access to the Jaeger Query API so you can pinpoint latency bottlenecks and failing spans quickly. ## Core Features & Use Cases - Service and Operation Discovery: List all traced services and their operations before diving into individual traces. - Latency and Error Analysis: Compute p50/p95/p99 latency statistics per operation, find slow traces above a duration threshold, and isolate traces containing error spans. - Full Trace Inspection: Retrieve a complete trace by ID and render its span tree with durations, tags, and error markers to identify the root-cause span. - Use Case: When users report slow checkout requests, run latency stats for the checkout service, pull the slowest traces, then inspect a specific trace to discover that a database span is the bottleneck. ## Quick Start Ask the agent to list the services in Jaeger and show latency statistics for the checkout service over the last hour.
