Regression Detector
Compare error rates before and after a deploy to detect regressions
Example Prompt
Compare error rates from the last deploy and let me know if there are any regressions
About
Monitors application health across deployments by pulling error data from Sentry and correlating it with deploy timestamps. Compares error frequency and new issue types pre- and post-deploy, then alerts the team via Slack if a regression is detected.
Workflow Steps
Query the deploy history from the database to identify the latest deploy timestamp
📊 Run SQL QueryFetch Sentry issues created before the deploy window for a baseline error rate
🚨 List IssuesFetch Sentry issues created after the deploy and compare volumes and new issue types
🚨 List IssuesAnalyze the delta between pre- and post-deploy error rates to classify severity
🧠 Structured ReasoningPost a regression report to the team Slack channel with findings and recommendations
📨 Send MessageTools Used
Run SQL Query
Execute a read-only SQL query against a PostgreSQL database and return results
Get Error Event
Retrieve detailed information about a specific Sentry error event including stack trace
List Issues
Retrieve unresolved error issues from a Sentry project, sorted by frequency or last seen
Structured Reasoning
Break down a complex problem into sequential reasoning steps with explicit chain-of-thought
Send Message
Send a message to a Slack channel or direct message conversation
Required MCP Servers
PostgreSQL
Query, inspect, and manage PostgreSQL databases with full SQL support
Sentry
Query error reports, performance data, and release health from Sentry
Sequential Thinking
Dynamic problem-solving through structured thought sequences with branching and revision
Slack
Send messages, read channels, and manage Slack workspace interactions