---

name: sentry-debugging description: Use when debugging errors with Sentry context, investigating production issues, triaging Sentry alerts, or when user mentions "sentry", "error tracking", "production error", or "investigate issue"

Debugging with Sentry

Guide for using Sentry data to debug and resolve production errors effectively.

Available Tools

Use these MCP tools to interact with Sentry:

Debugging Workflow

1. Identify the Issue

Start by listing recent issues:

CopyUse mcp__sentry__list_issues with:
- organization: (from settings or ask user)
- status: "unresolved"
- limit: 10

2. Get Issue Details

Once you have an issue ID, get full details:

CopyUse mcp__sentry__get_issue with:
- issueId: "PROJ-123" or numeric ID

Key information to extract:

• metadata.errorType: The exception type (TypeError, ValueError, etc.)
• metadata.errorValue: The error message
• metadata.filename: File where error occurred
• metadata.function: Function name
• topTags: Browser, OS, environment, release info

3. Analyze Events

Look at individual occurrences for patterns:

CopyUse mcp__sentry__list_issue_events with:
- issueId: the issue ID
- limit: 5

Check for:

• Common user patterns
• Browser/device patterns
• Time patterns (when do errors spike?)

4. Find Related Code

Using the filename and function from issue metadata:

1. Search the codebase for the file
2. Read the relevant function
3. Look for potential causes based on the error type

5. Fix and Verify

After implementing a fix:

1. Deploy the change
2. Monitor Sentry for new occurrences
3. Resolve the issue: mcp__sentry__update_issue_status with status "resolved"

Common Error Patterns

TypeError: Cannot read property 'X' of undefined

• Check for missing null checks
• Verify data is loaded before accessing
• Look for race conditions in async code

Network Errors

• Check API endpoint availability
• Verify authentication tokens
• Look for CORS issues
• Check timeout settings

Database Errors

• Check connection pool settings
• Look for transaction deadlocks
• Verify query parameters

Settings

Configure defaults in .claude/sentry.local.md:

Copy---
organization: your-org-slug
project: your-project-slug
---

Optional notes about your Sentry setup.

Quick Commands

• /sentry:issues - List recent issues
• /sentry:issue <id> - Get issue details
• /sentry:resolve <id> - Mark issue as resolved
• /sentry:stats - View project error statistics
• /sentry:find-event <event-id> - Find issue by event ID

Tips

1. Prioritize by user impact: Sort issues by userCount to fix what affects most users
2. Check release tags: New issues after a release often indicate regressions
3. Use environment tags: Filter by production vs staging to focus on real issues
4. Link issues to code: When you find the fix, note the commit in Sentry