• TRIAGE
  • CONTEXT
  • REPAIR
  • HUMAN

How to debug production errors with your AI

Connect Sentry to your AI client once. From then on, your agent handles the heavy lifting of digging through error logs and metadata.

Ask AI about this page

Short answer

How can my AI help me debug Sentry errors?

Your AI pulls stack traces and environmental metadata from Sentry to find the exact line of code causing a crash. You get a clear path to the fix through this breakdown.

The outcomes.

Where your debugging efforts land.

The AI categorizes and processes incoming error data into actionable outcomes.

TRIAGE

Identified root causes

The AI examines the error context to tell you exactly what broke. You get the specific file and line number immediately.

CONTEXT

Environmental snapshots

Your agent gathers the specific OS, browser, or device info tied to the crash. This helps you replicate the bug in your local environment.

REPAIR

Suggested code fixes

By looking at the stack trace, your AI proposes the logic changes needed to stop the error. You can then review these against your codebase.

HUMAN

Final code deployment

The AI cannot push code to your repository or deploy fixes. You remain in control of the actual merge and release process.

The workflow

What your AI does when an error occurs.

The AI automates the investigation phase of the debugging lifecycle.

  1. Fetch error details

    The AI grabs the core error message and the specific stack trace from the reported issue.

    get_issue_details

  2. Inspect event context

    It looks deeper into the specific event to see the exact state of the environment when the crash happened.

    get_event_details

  3. Analyze the trace

    The agent maps the stack trace to your logic to find the broken line.

    get_issue_details

  4. Summarize findings

    The AI compiles the error, the environment, and the code location into a single report for you.

    get_event_details

Try it

Copy these to start.

Use these prompts to kick off an investigation.

Starting points

These are opening lines. Replace the bracketed text with your specific Sentry issue IDs or project names.

Sentry Application Monitoring Connector

You're all set. Choose your MCP client and follow the setup instructions.

Connector linkhttps://edge.vinkius.com/vk_preview_44oqnI7Ime1XiKScWG9YlmefdW04ekDjWjnlXHRM/mcp

Claude Desktop

Follow the steps below to connect in seconds.

  1. 1In Claude Desktop, open Settings → Connectors.
  2. 2Click “Add custom connector” and paste the connector link above as the remote MCP server URL.
  3. 3Click Add and start a new chat — Sentry Application Monitoring capabilities are ready to use.
Configuration · claude_desktop_config.jsonCopy
{
  "mcpServers": {
    "sentry-mcp": {
      "url": "https://edge.vinkius.com/vk_preview_44oqnI7Ime1XiKScWG9YlmefdW04ekDjWjnlXHRM/mcp"
    }
  }
}
Copy into chat04

  • Look up the details for Sentry issue [ID] and tell me why it's failing.

  • Check the environmental metadata for event [ID] to see if it's a specific browser issue.

  • Analyze the stack trace for the latest error in my [Project Name] project.

  • Summarize the root cause for issue [ID] based on the available stack trace.

  • Claude
  • ChatGPT
  • Cursor
  • VS Code
  • Windsurf
  • Claude Code
  • JetBrains
  • Cline

Start here

Connect Sentry once, then ask.

Just link your Sentry account. Your credentials stay encrypted, and you can start debugging immediately in your preferred AI client.

Connect Sentry to your AI

FAQ

How the debugging process works

  • 01

    Can the AI fix the code automatically?

    No. The AI identifies the error and suggests a fix, but you must manually apply and test the code.

  • 02

    Can the AI delete issues in Sentry?

    No. The tools only allow the AI to read issue and event details. It cannot modify or delete anything in your Sentry account.

  • 03

    Does the AI see my entire codebase?

    No. It only sees the specific stack traces and metadata provided by Sentry for the issues you ask about.

  • 04

    Can I use this with Cursor or VS Code?

    Yes. Any MCP-compatible client like Cursor, VS Code, or Claude can use this once connected.

  • 05

    What if an issue doesn't have a stack trace?

    The AI will use the available event details to try and infer the cause, but it won't be able to point to a specific line of code without a trace.

  • More questions about Sentry? The Connector page answers them. See everything the Sentry Connector can do

Connect Sentry Application Monitoring to Claude, Cursor, ChatGPT & more