# Sentry MCP for AI Agents AI Agent Connect

> Sentry MCP gives your AI agent direct access to Sentry's application performance monitoring. It lets your agent pull raw exceptions, read stack traces, and manage issue statuses in real time. Use it to triage production crashes and resolve bugs directly from your AI client.

## Overview
- **Category:** loved-by-devs
- **Price:** Free
- **Endpoint:** https://edge.vinkius.com/vk_preview_44oqnI7Ime1XiKScWG9YlmefdW04ekDjWjnlXHRM/ai-agent-connect
- **Tags:** error-tracking, performance-monitoring, debugging, stack-trace, incident-management

## Description

The Sentry MCP gives your AI agent direct access to Sentry's application performance monitoring. You spend way too much time switching tabs between your code and the Sentry dashboard just to figure out why a production build is melting down. This Connector changes that by letting your AI agent do the heavy lifting for you. Instead of hunting for a specific error ID, you can just tell your agent to find the latest issues in your project and summarize what's actually happening. It pulls the stack traces directly into your chat window so you can see the exact line of code that failed without leaving your workspace. It's a huge time saver when you're trying to catch a bug before it hits your users. You can even have your agent handle the boring stuff, like closing out issues once you've pushed a fix. This is part of the Vinkius catalog, which makes it easy to plug into your existing workflow with just a few credentials. It gives you a direct line into your operational data so you can stop guessing and start fixing.

## Tools

### delete_issue
The delete_issue tool permanently removes an issue from your project. Use it only for cleanup because this action is irreversible.

### list_events
The list_events tool lists recent events for a specific project. It helps you see the chronological flow of what's happening in your app.

### list_organization_users
The list_organization_users tool lists all users in an organization. Use this to check who has access to your Sentry environment.

### get_event_details
The get_event_details tool pulls the full details for a single event. It's the best way to see the specific metadata behind a single occurrence.

### list_projects
The list_projects tool gets a list of all projects within an organization. Use this to scope your AI's search to a specific app or service.

### resolve_issue
The resolve_issue tool marks an issue as resolved in Sentry. This helps keep your dashboard clean and signals that a bug is fixed.

### list_organization_teams
The list_organization_teams tool lists all teams associated with an organization. This is useful for auditing permissions or looking up team structures.

### list_organizations
The list_organizations tool shows all the organizations you have access to. This helps you navigate between different workspace environments.

### get_issue_details
The get_issue_details tool pulls the full details and stack trace for a specific issue. This is the primary way to let your agent analyze a crash.

### list_issues
The list_issues tool lists all errors and issues in a project. Use this to find what's currently breaking in your production environment.

## Prompt Examples

**Prompt:** 
```
Why did my app crash in the last hour?
```

**Response:** 
```
I've checked your recent logs. Your app experienced 12 crashes in the last 60 minutes.

**Primary Issue:**
- **Error:** `TypeError: Cannot read properties of undefined (reading 'auth_token')`
- **Frequency:** 8 occurrences
- **Impact:** Affecting the login flow.

Would you like me to pull the full stack trace for this specific error?
```

**Prompt:** 
```
Get details on issue 6B3VX4921.
```

**Response:** 
```
Here are the details for issue **6B3VX4921**:

- **Type:** PostgreSQL Deadlock
- **Location:** `payment_processor.ts` at Line 132
- **Context:** This occurred during a heavy write operation.
- **Impact:** 300+ users affected since Tuesday.

It looks like this is tied to the recent migration you pushed.
```

**Prompt:** 
```
Close the bug for the login page.
```

**Response:** 
```
I've successfully updated the status of the login page error in Sentry. It is now marked as **RESOLVED**. I've also updated your dashboard to reflect this change.
```

## Capabilities

### Fetch latest crash logs
The agent pulls the most recent errors from your project to show you what is breaking right now.

### Analyze stack traces
Your AI can read the full technical trace of a crash to pinpoint the exact line of code at fault.

### Close resolved bugs
You can tell your agent to mark issues as resolved without having to manually update the dashboard.

### Scan project structures
The agent can list all projects and organizations to help you navigate different environments.

### Audit team memberships
You can check who has access to your Sentry workspace or see how teams are organized.

### Monitor live errors
The agent can pull specific event details to help you see the context behind a single failure.

## Use Cases

### Production Crash Triage
A deployment causes a spike in errors. An engineer asks the agent to find the most frequent crash and summarize the stack trace using `list_issues` and `get_issue_details`.

### Database Deadlock Investigation
A specific error ID is flagged. The user asks the agent to fetch all metadata for that issue to see if it correlates with a recent database migration.

### Workspace Security Audit
A founder wants to know who has access to the production environment. They ask the agent to list all users and teams in the organization.

### Daily Standup Summary
Before a meeting, a lead asks the agent to list all high-priority issues across all projects to get a quick status update on the day's bugs.

## Benefits

- You can triage production crashes faster by using `list_issues` to see what is breaking immediately.
- You get deep context on bugs because `get_issue_details` feeds the exact stack trace to your agent.
- You keep your dashboard clean by using `resolve_issue` to mark fixed bugs without opening the web UI.
- You can audit your workspace security by using `list_organization_users` to see who has access.
- You can manage multiple apps easily by using `list_projects` to switch contexts for different services.
- You get a clear history of app behavior by using `list_events` to see the sequence of actions leading to a crash.

## How It Works

The bottom line is you get instant access to Sentry's error data without leaving your IDE.

1. Connect the Sentry MCP to your Vinkius dashboard.
2. Provide your Organization Slug and a scoped Auth Token.
3. Ask your agent to summarize recent crashes or find specific error details.

## Frequently Asked Questions

**How does the Sentry MCP help me debug faster?**
It allows your AI agent to pull real-time error logs and stack traces directly into your chat. You can ask questions about specific crashes and get immediate answers without leaving your workspace.

**Can I use the Sentry MCP to manage my issues?**
Yes. You can tell your AI agent to mark issues as resolved or find specific ones in your project. It handles the manual clicking for you.

**Does the Sentry MCP work with my current IDE?**
Yes. Since it's an Connector, it works with any compatible client like Cursor, Windsurf, or VS Code, giving you Sentry data right where you write code.

**Is my Sentry data secure with this Connector?**
The Connector uses your own scoped Auth Token and Organization Slug. You control exactly what parts of your Sentry environment the AI can see.

**Can the AI agent see my whole Sentry account?**
It depends on the permissions of the Auth Token you provide. You can scope it to specific projects or organizations to keep it limited.

**What happens when I ask about a crash?**
The agent will search your project for recent errors, find the most relevant ones, and summarize the technical details and stack traces for you.

**Can this AI integration actually mark errors as fixed?**
Yes. This agent component possesses mutable write access. If you invoke the prompt properly, it will fire the `resolve_issue` tool, marking the corresponding exception ID completely dealt with inside the Sentry ecosystem. It can also erase bugs fully via `delete_issue`.

**What is the difference between inspecting an 'Issue' and an 'Event'?**
An 'Issue' gathers underlying multiple occurrences of the identical stack exception into one overarching master group. In contrast, querying an 'Event' (`get_event_details`) focuses the AI on a strictly singular, point-in-time incidence where the system crashed.

**Do I need to supply the Organization Slug with every command?**
You configure the overarching Organization Slug strictly at startup globally. For project-level filters, let your LLM query `list_projects` first to fetch internal slugs naturally into its memory buffer without constant repeated user inputs.