What problem does it solve? When users report that something is broken or error rates spike, teams need to quickly identify what failed, which users are affected, and why. This Skill triages errors across Amplitude's three auto-captured error events to surface root causes instead of isolated symptoms. ## Core Features & Use Cases - Cross-Event Correlation: Links failed network requests, JavaScript errors, and error clicks by page path to reveal causal chains such as a 500 response triggering a TypeError that users then click on. - Impact Quantification: Counts affected users, segments errors by platform or plan tier, and classifies severity from Critical to Low based on user impact percentages. - Root Cause Hypothesis: Correlates error spikes with deployments, experiments, and customer feedback, then attaches session replay links for visual confirmation. - Use Case: After a deploy, ask what broke. The Skill compares pre- and post-deploy error volumes, identifies new error messages, maps them to failing API endpoints, and delivers a triage report with replay links and recommended fixes. ## Quick Start Ask your AI assistant to investigate what's broken in your Amplitude project over the last 7 days and rank the top errors by user impact.
