Files
gh-coalesce-labs-catalyst-p…/commands/debug_production_error.md
2025-11-29 18:14:36 +08:00

4.1 KiB

description, category, tools, model, version
description category tools model version
Debug production errors using Sentry error tracking and analysis debugging Task, TodoWrite inherit 1.0.0

Debug Production Error

Investigate production errors using Sentry's error tracking, stack traces, and context.

Prerequisites

  • Sentry MCP must be enabled (this plugin should be enabled)
  • Environment variables configured:
    • SENTRY_AUTH_TOKEN
    • SENTRY_ORG
    • SENTRY_PROJECT

Usage

/catalyst-dev:debug-production-error <error-description-or-id>

Examples:
  /catalyst-dev:debug-production-error "TypeError in checkout flow"
  /catalyst-dev:debug-production-error "ISSUE-123"
  /catalyst-dev:debug-production-error "errors from last deployment"
  /catalyst-dev:debug-production-error "unhandled exceptions this week"

What This Command Does

Uses Sentry MCP tools to:

  1. Search for relevant errors
  2. Retrieve stack traces and context
  3. Analyze error patterns and frequency
  4. Identify affected users and environments
  5. Suggest root causes and fixes

Available Sentry Capabilities

When this plugin is enabled, you have access to ~19 Sentry tools:

Error Search & Analysis:

  • Search issues by query
  • Filter by status, assignment, date
  • View error trends and patterns
  • Identify new vs recurring errors

Stack Trace Analysis:

  • Full stack traces with source context
  • Source map resolution
  • Frame-by-frame analysis
  • Variable inspection

Context & Metadata:

  • User context (who was affected)
  • Environment details
  • Release/deployment information
  • Breadcrumb trail (user actions leading to error)

Issue Management:

  • Update issue status
  • Assign to team members
  • Link to tickets/PRs
  • Add comments and notes

Root Cause Analysis (Seer AI):

  • AI-powered root cause identification
  • Code-level explanations
  • Specific fix recommendations
  • Related error patterns

Example Debugging Sessions

Investigate Specific Error

/catalyst-dev:debug-production-error "Show me details for MYAPP-456 including stack trace and user context"

Search by Error Type

/catalyst-dev:debug-production-error "Find all TypeError exceptions in the last 24 hours"

Deployment Issues

/catalyst-dev:debug-production-error "What new errors appeared after release v2.3.0?"

High-Impact Errors

/catalyst-dev:debug-production-error "Show unresolved errors affecting more than 100 users"

Output Format

Analysis typically includes:

Error Overview:

  • Error message and type
  • Frequency and trend
  • First seen / last seen
  • Number of users affected

Stack Trace:

  • Full call stack
  • Source code context
  • File paths and line numbers
  • Variable values (if available)

User Context:

  • User ID and properties
  • Browser/device information
  • URL and user actions (breadcrumbs)

Root Cause (when Seer analysis available):

  • Likely cause explanation
  • Relevant code snippets
  • Specific fix recommendations
  • Related issues

Advanced Queries

Filter by Environment

/catalyst-dev:debug-production-error "production errors in payment service"

Time-Based Analysis

/catalyst-dev:debug-production-error "spike in errors between 2pm-3pm today"

User-Specific

/catalyst-dev:debug-production-error "errors for user@example.com"

Integration with Analytics

If you have both plugins enabled:

# Enable both
/plugin enable catalyst-debugging
/plugin enable catalyst-analytics

# Combined analysis
> "Show me errors in checkout AND how many users abandoned checkout today"

Workflow Integration

With Issue Tracking

After identifying root cause:

> "Create a GitHub issue for this error with the stack trace and fix recommendations"

With Code Changes

After finding the bug:

/create-plan "Fix the TypeError in checkout.ts based on Sentry analysis"

Context Cost

This plugin adds ~20,670 tokens to your context window. Disable when debugging is complete:

/plugin disable catalyst-debugging

See also: /error-impact-analysis, /trace-analysis