Features

MCP Server

Model Context Protocol (MCP) lets AI assistants interact with your DeadPing account. Ask your AI to check monitor status, investigate incidents, or search through job output – all in natural language.

The DeadPing MCP server lets AI assistants like Claude, Cursor, and Windsurf manage your monitors through Model Context Protocol. Install it as an npm package or connect to the hosted remote server, and your AI assistant can create monitors, check status, search output, and investigate incidents using natural language.

What Is MCP?

Model Context Protocol is an open standard that allows AI assistants to connect to external tools and data sources. DeadPing exposes a remote MCP server that any compatible client can connect to.

Setup

The DeadPing MCP server is hosted at https://deadping.io/api/mcp and authenticates via your API key.

Claude Desktop

Add the following to your claude_desktop_config.json:

claude_desktop_config.json

{
  "mcpServers": {
    "deadping": {
      "url": "https://deadping.io/api/mcp",
      "headers": {
        "X-API-Key": "dp_live_your_key_here"
      }
    }
  }
}

Claude Code

Add DeadPing as a remote MCP server in your Claude Code settings or .mcp.json:

.mcp.json

{
  "mcpServers": {
    "deadping": {
      "url": "https://deadping.io/api/mcp",
      "headers": {
        "X-API-Key": "dp_live_your_key_here"
      }
    }
  }
}

Cursor / Windsurf

Both Cursor and Windsurf support MCP servers. Add the DeadPing server URL and API key in your editor's MCP settings. The configuration format is the same as above.

Authentication

All requests to the MCP endpoint require an X-API-Key header with a valid DeadPing API key. Create one at Dashboard → API Keys or via the API Keys documentation.

Available Tools

The MCP server exposes the following tools that AI assistants can call:

Tool	Description
`list_monitors`	List all monitors with current status
`get_monitor`	Get details for a specific monitor including recent pings
`create_monitor`	Create a new monitor with name, schedule, and grace period
`pause_monitor`	Pause a monitor (stops alerting)
`resume_monitor`	Resume a paused monitor
`delete_monitor`	Delete a monitor and all its ping history
`list_incidents`	List recent incidents with status and affected monitors
`get_incident`	Get full incident detail including timeline and postmortem
`update_incident`	Add root cause and action items to a resolved incident
`search_output`	Full-text search across captured ping output (Pro+)
`list_anomalies`	List recent anomaly events, filter by monitor or type (Pro+)
`get_monitor_baseline`	View statistical baseline (mean, stddev, sample size) for a monitor (Pro+)
`get_status`	Get account overview: total monitors, active incidents, uptime stats

Example Interactions

Once connected, you can interact with DeadPing using natural language. Here are some examples:

Check monitor status

prompt

"Which monitors are currently down?"

→ The AI calls list_monitors and filters by status.
  Response: "2 monitors are down: 'nightly-backup' (missed 3 hours ago)
  and 'db-cleanup' (missed 45 minutes ago)."

Manage monitors

prompt

"Pause the ETL Pipeline monitor"

→ The AI calls list_monitors to find the ID, then calls pause_monitor.
  Response: "Done. 'ETL Pipeline' is now paused and won't trigger alerts."

Investigate incidents

prompt

"Show me the latest incident"

→ The AI calls list_incidents to get the most recent, then get_incident for details.
  Response: "Incident #42 started 2 hours ago. 5 monitors affected:
  backup-db, sync-users, generate-reports, cleanup-logs, health-check.
  All recovered after 18 minutes. No postmortem added yet."

Check anomalies

prompt

"Are there any anomalies on my monitors?"

→ The AI calls list_anomalies to check recent events.
  Response: "1 anomaly detected: 'billing-sync' took 4x longer than
  usual (12.3s vs baseline 3.1s). Flagged as critical 2 hours ago."

Search output

prompt

"Search for errors in the backup output"

→ The AI calls search_output with query "error" filtered to backup monitors.
  Response: "Found 3 matches in the last 24 hours:
  - 'ERROR: connection refused to database server' (6h ago)
  - 'ERROR: disk space below threshold' (18h ago)
  - 'ERROR: timeout waiting for lock' (23h ago)"

Compatible Clients

The DeadPing MCP server works with any client that supports the Model Context Protocol, including:

Claude Desktop – Anthropic's desktop app
Claude Code – Anthropic's CLI coding agent
Cursor – AI-powered code editor
Windsurf – AI-powered code editor by Codeium
Any other MCP-compatible client