Skip to content

sentry.mdx

Latest commit

 

History

History
195 lines (144 loc) · 7.67 KB

sentry.mdx

File metadata and controls

195 lines (144 loc) · 7.67 KB
Raw
title Sentry Integration
sidebarTitle Sentry
icon sentry

import { COMMUNITY_SLACK_URL } from "/snippets/links.mdx";

Integrate Codegen with your Sentry workspace to allow agents to analyze errors, manage issues, and provide intelligent debugging assistance.

{/* TODO: Add Sentry integration screenshot */}

The Sentry integration is currently in beta. Please reach out in the community to have it enabled for your Codegen account.

Overview

The Sentry integration enables Codegen agents to:

  • Analyze Production Errors: Get detailed insights into errors and exceptions happening in your applications
  • Root Cause Analysis: Use AI-powered analysis to understand why errors are occurring
  • Issue Management: Update issue status, assign issues, and add comments
  • Project Management: Create and manage Sentry projects and teams
  • Release Tracking: Monitor releases and their associated issues

Step-by-Step Setup Guide

Follow these steps to successfully connect Codegen to your Sentry workspace:

1. Enable Sentry Integration for Your Organization

The Sentry integration uses the Model Context Protocol (MCP) to provide secure access to your Sentry data. Contact the Codegen team to enable this feature for your organization.

2. Connect Your Sentry Account

  • Go to Codegen's Sentry Integration page.
  • Click Connect Sentry Workspace and complete the OAuth flow.
  • Grant the necessary permissions for Codegen to access your Sentry data.
Make sure you have the appropriate permissions in your Sentry organization to authorize third-party applications.

3. Configure Permissions

The Sentry integration requires the following OAuth scopes:

  • org:read - Read organization information
  • project:write - Create and manage projects
  • team:write - Create and manage teams
  • event:write - Access error events and issues

4. Start Using Codegen with Sentry

Once connected, you can:

  • Ask Codegen to analyze specific Sentry issues by providing issue URLs
  • Request error analysis and debugging suggestions
  • Have Codegen manage issue status and assignments
  • Get insights into error patterns and trends

Capabilities

The Sentry integration provides comprehensive access to your Sentry workspace:

Error Analysis & Debugging

  • Issue Details: Fetch comprehensive information about specific errors and exceptions
  • Stacktrace Analysis: Get detailed stacktraces and error context
  • Root Cause Analysis: AI-powered analysis to identify the underlying causes of issues
  • Error Patterns: Identify recurring issues and patterns across your applications

Issue Management

  • Status Updates: Change issue status (resolved, ignored, unresolved)
  • Assignment: Assign issues to team members or teams
  • Comments: Add comments and updates to issues
  • Bulk Operations: Manage multiple issues efficiently

Project & Team Management

  • Project Creation: Create new Sentry projects for your applications
  • Team Management: Create and manage teams within your organization
  • DSN Management: Create and manage Data Source Names (DSNs) for your projects

Monitoring & Insights

  • Release Tracking: Monitor releases and their associated issues
  • Error Trends: Analyze error frequency and patterns over time
  • Performance Insights: Get insights into application performance issues

How Agents Use Sentry

Agents leverage the Sentry integration to provide intelligent debugging assistance:

Error Investigation

Human: "Can you analyze this Sentry issue: https://my-org.sentry.io/issues/PROJECT-123"

Agent: I'll analyze that Sentry issue for you. Let me fetch the details and provide insights...

Proactive Issue Management

  • Automatic Triage: Agents can automatically categorize and prioritize issues
  • Status Updates: Keep issue status current based on code changes and fixes
  • Team Notifications: Notify relevant team members about critical issues

Code-Error Correlation

  • Link PRs to Issues: Connect GitHub pull requests to related Sentry issues
  • Fix Validation: Verify that code changes resolve reported issues
  • Regression Detection: Identify when new code introduces previously resolved issues

Technical Implementation

The Sentry integration is built using:

Model Context Protocol (MCP)

  • Secure Communication: Uses MCP for secure, standardized communication with Sentry
  • Real-time Access: Direct connection to Sentry's API through MCP server
  • OAuth Authentication: Secure token-based authentication with automatic refresh

Backend Architecture

  • Token Management: Automatic OAuth token refresh and management
  • Feature Flags: Organization-level feature flag control
  • Error Handling: Robust error handling and timeout management

Available Tools

When the integration is enabled, agents have access to Sentry-specific tools:

  • sentry_whoami - Get current user information
  • sentry_find_organizations - List accessible organizations
  • sentry_find_projects - List projects in an organization
  • sentry_find_issues - Search for issues with advanced filtering
  • sentry_get_issue_details - Get comprehensive issue information
  • sentry_update_issue - Update issue status and assignments
  • sentry_analyze_issue_with_seer - AI-powered root cause analysis

Best Practices

For Development Teams

  1. Consistent Naming: Use clear, descriptive names for projects and teams
  2. Regular Monitoring: Set up agents to regularly check for new critical issues
  3. Integration Workflow: Incorporate Sentry analysis into your PR review process

For Issue Management

  1. Automated Triage: Let agents handle initial issue categorization
  2. Status Tracking: Keep issue status updated as fixes are deployed
  3. Documentation: Use agent-generated analysis to document complex issues

Security Considerations

  1. Token Security: OAuth tokens are securely managed and automatically refreshed
  2. Scope Limitation: Only request necessary permissions for your use case
  3. Access Control: Use Sentry's built-in access controls to limit data exposure

Troubleshooting

Common Issues

Integration Not Available

  • Ensure the Sentry MCP integration is enabled for your organization
  • Contact support if you don't see the integration option

Authentication Errors

  • Check that your OAuth token hasn't expired
  • Verify you have the necessary permissions in your Sentry organization
  • Try disconnecting and reconnecting the integration

Timeout Issues

  • The integration has built-in timeout handling for API calls
  • Large organizations may experience longer response times

Getting Help

If you encounter issues with the Sentry integration:

  1. Check the troubleshooting guide
  2. Join our community Slack for support
  3. Contact the Codegen team for organization-specific issues

Example Usage

Here are some common ways to use the Sentry integration:

Analyzing a Specific Issue

@codegen Can you analyze this Sentry issue and suggest a fix?
https://my-org.sentry.io/issues/PROJECT-123

Bulk Issue Management

@codegen Please review all unresolved issues in the backend project 
and prioritize them by frequency and impact

Release Monitoring

@codegen Check if the latest release introduced any new errors 
and create GitHub issues for any critical problems

The Sentry integration makes error monitoring and debugging more efficient by providing AI-powered analysis and automated issue management capabilities.