Skip to content

Repository files navigation

elmah.io AI Plugin

Error monitoring and log management for your production apps, directly from your AI editor.

This plugin connects any MCP-compatible AI editor to the elmah.io MCP server, giving you and your AI agent direct access to production logs, error details, deployment history, and uptime monitors — without leaving your editor.

Supported Editors

Editor Manifest
Cursor .cursor-plugin/
Claude Code .claude-plugin/
Any MCP-compatible editor Point at https://mcp.elmah.io/mcp

Installation

Cursor

OAuth (recommended):

{
  "mcpServers": {
    "elmah.io": {
      "url": "https://mcp.elmah.io/mcp",
      "auth": {
        "CLIENT_ID": "ai-plugin"
      }
    }
  }
}

API key:

{
  "mcpServers": {
    "elmah.io": {
      "url": "https://mcp.elmah.io/mcp",
      "headers": {
        "Authorization": "Bearer YOUR_API_KEY"
      }
    }
  }
}

Claude Code

OAuth (recommended):

{
  "mcpServers": {
    "elmah.io": {
      "type": "http",
      "url": "https://mcp.elmah.io/mcp",
      "oauth": {
        "clientId": "ai-plugin"
      }
    }
  }
}

On first use, your editor will prompt you to sign in with your elmah.io account. Access tokens expire after 30 days; re-authenticate if you see auth errors.

API key:

{
  "mcpServers": {
    "elmah.io": {
      "type": "http",
      "url": "https://mcp.elmah.io/mcp",
      "headers": {
        "Authorization": "Bearer YOUR_API_KEY"
      }
    }
  }
}

Skills

Skill Description
debug-production-error Investigate a specific error by ID, message, or stack trace
post-deployment-check Verify production health after shipping code
investigate-frequent-errors Find and prioritize your highest-impact errors
check-uptime-status Check monitor status and investigate downtime events

Usage examples

Debug this production error: NullReferenceException in OrderService
Run a post-deployment check after my latest deploy
What are the most frequent errors in my production log this week?
Is my site up? Check uptime status

Agents

deployment-watchdog — A background agent that monitors for regressions for 30 minutes after a deployment. Alerts immediately on fatal errors or downtime; otherwise reports back with a final HEALTHY / UNHEALTHY verdict.

Start the deployment watchdog for version 2.4.1

MCP Tools Available

Once connected, the following elmah.io tools are available to your AI agent:

  • organizations_list / organizations_get_details — list organizations and get quota/stats
  • logs_list / logs_get_details — list logs and get configuration details
  • messages_list_recent — fetch recent log messages with filtering
  • messages_list_frequent — get the most frequently occurring errors
  • messages_get — fetch a single message with full details (stack trace, custom data, etc.)
  • messages_count — count messages matching a query
  • deployments_list — list deployments with timestamps and versions
  • uptime_list / uptime_get_details — list uptime monitors and get incident history
  • heartbeats_list / heartbeats_get_details — list heartbeat monitors and check health
  • users_get_current — get the authenticated user
  • system_get_context — get system information like current date and time

Requirements

  • An active elmah.io account
  • At least one log configured in elmah.io

License

Apache 2.0 — see LICENSE

Contributors

Languages