What problem does it solve?
Deciding whether a new software release is safe to continue rolling out is hard when error data is noisy, rollouts are incomplete, or instrumentation is missing. This Skill turns raw Sentry signals into an evidence-based release health verdict so teams avoid both false alarms and missed regressions.
Core Features & Use Cases
- Release Comparison: Compares a target release against the previous stable release over equivalent time windows in the production environment.
- Regression Detection: Reports new and regressed issues, affected user counts, severity, and high-confidence change points.
- Evidence-Based Recommendations: Distinguishes incomplete rollouts, low traffic, and missing instrumentation from genuinely healthy behavior, and only recommends continue, monitor, pause, or rollback when the evidence supports it.
- Use Case: After deploying version 2.4.0 to 20% of users, ask the agent to compare its Sentry error rates and regressions against 2.3.x and receive a grounded recommendation on whether to continue the rollout.
Quick Start
Ask the agent to assess the release health of version 2.4.0 in production using Sentry data and compare it with the previous stable release.