Sentry Issues

🚀 Sentry Issue MCP Project

This is an MCP (Middleware Handler) for processing Sentry issues. It supports two tools to retrieve a single issue or a list of issues, and you can have a Large Language Model (LLM) analyze the results or perform other operations.

🚀 Quick Start

Here is a configuration example for the MCP server:

  "mcpServers": {
    "sentry-issue-mcp": {
      "type": "stdio",
      "command": "npx",
      "args": [
        "-y",
        "sentry-issues-mcp@latest"
      ],
      "env": {
        "SENTRY_HOST": "<your_sentry_host>",
        "SENTRY_ORG": "<your_sentry_org>",
        "SENTRY_PROJ": "<your_sentry_proj>",
        "SENTRY_USER_TOKEN": "<your_sentry_user_token>"
      }
    }
  }

✨ Features

• Compact Size: It has a small footprint.
• Easy to Understand: The codebase is straightforward.
• User-Friendly: Simple to use.

📦 Tool Explanation

get_single_event

• Retrieve detailed information about a specific event. In the tiny mode, it returns stack information; in the huge mode, it returns all information.
• Input Parameters:
  • url_or_id: Sentry event link or Sentry event ID.
  • organization_id_or_slug: Sentry organization ID or slug (optional).
  • project_id_or_slug: Sentry project ID or slug (optional).
  • mode: tiny or huge (optional).

get_project_events

• Retrieve a list of events for a specific project. In the tiny mode, it returns IDs and titles; in the huge mode, it returns all information.
• Input Parameters:
  • project_id_or_slug: Sentry project ID or slug.
  • organization_id_or_slug: Sentry organization ID or slug (optional).
  • mode: tiny or huge (optional).

💻 Usage Examples

Basic Usage

Let the LLM analyze an issue (via URL or ID)

1. Input: "Analyze the issue in this link and provide the cause and how to fix it, {sentry_issue_url}"
2. If your LLM is intelligent enough, it will call the tool for processing.
3. You will get the results.

Let the LLM find the most critical issue today (default time range is "24 hours")

1. Input: "Find the most critical issue today and provide the cause and how to fix it."
2. If your LLM is intelligent enough, it will call the tool for processing.
3. You will get the results.

📄 License

MIT

⚠️ Important Note

Starting from version 1.0.5, the function to retrieve issues has been deprecated. It was considered redundant as its response structure was very similar to the event API.