Pulsetic MCP Server

Connect AI assistants to your Pulsetic monitoring platform using the Model Context Protocol (MCP).

What is MCP?

The Model Context Protocol is an open standard that allows AI assistants to interact with external tools and services. With Pulsetic's MCP server, you can ask your AI assistant to check monitor status, create incidents, schedule maintenance, and more — all through natural conversation.

Supported clients:

• Claude Desktop — Anthropic's desktop app
• Claude Code — Anthropic's CLI for developers
• ChatGPT — OpenAI's AI assistant (via OAuth)
• Cursor — AI-powered code editor
• Windsurf — AI-powered IDE
• Any client implementing the MCP specification

Available Tools

The MCP server exposes 17 tools organized into four categories.

Monitors

Status Pages

Incidents

Maintenance

Testing with cURL

You can test the MCP endpoint directly with cURL to verify your API key works.

Initialize

curl -X POST https://api.pulsetic.com/mcp \
  -H "Authorization: YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "jsonrpc": "2.0",
    "id": 1,
    "method": "initialize",
    "params": {
      "protocolVersion": "2025-03-26",
      "capabilities": {},
      "clientInfo": {"name": "curl", "version": "1.0"}
    }
  }'

List tools

curl -X POST https://api.pulsetic.com/mcp \
  -H "Authorization: YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc":"2.0","id":2,"method":"tools/list"}'

Call a tool

curl -X POST https://api.pulsetic.com/mcp \
  -H "Authorization: YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "jsonrpc": "2.0",
    "id": 3,
    "method": "tools/call",
    "params": {
      "name": "list_monitors",
      "arguments": {}
    }
  }'

Tool Reference

list_monitors

Returns all monitors for your account.

No parameters.

get_monitor

Returns full details for a single monitor, including SSL certificate info.

get_monitor_stats

Returns uptime %, downtime, incident count, and average response time across 1-day, 7-day, 30-day, and 90-day periods.

get_monitor_events

Returns online/offline events in a date range (max 100, newest first).

get_monitor_downtime

Returns total downtime in seconds and incident count for a period.

create_monitor

Creates a new uptime monitor.

update_monitor

Updates an existing monitor's settings. Only include the fields you want to change.

delete_monitor

Permanently deletes a monitor and all associated data.

pause_monitor

Pauses a monitor. It will stop being checked until resumed.

resume_monitor

Resumes a paused monitor.

list_status_pages

Returns all status pages for your account.

No parameters.

list_incidents

Returns incidents for a status page, including the latest update for each.

create_incident

Creates a new incident with an initial status update that appears on the status page.

update_incident

Posts a new status update to an existing incident.

resolve_incident

Marks an incident as resolved with a resolution message.

create_maintenance

Schedules a maintenance window on a status page.

delete_maintenance

Deletes a scheduled maintenance window.

API Reference

The MCP server implements the MCP specification (2025-03-26) using Streamable HTTP transport over a single POST /mcp endpoint with JSON-RPC 2.0.

Prerequisites

• A Pulsetic account on a plan with API access (Business or Enterprise).
• A Pulsetic API key — generate one from Settings → API in your dashboard.
• An MCP-compatible AI client.

Setup

Claude Desktop

Claude Desktop uses stdio-based MCP servers. Use mcp-remote to bridge the remote HTTP endpoint. Open Settings → Developer → Edit Config and add:

{
  "mcpServers": {
    "pulsetic": {
      "command": "npx",
      "args": [
        "-y", "mcp-remote",
        "https://api.pulsetic.com/mcp",
        "--header", "Authorization: Bearer YOUR_API_KEY"
      ]
    }
  }
}

Replace YOUR_API_KEY with your actual API key. Restart Claude Desktop to apply.

Claude Code

Add the server to your project's .mcp.json file:

{
  "mcpServers": {
    "pulsetic": {
      "type": "streamableHttp",
      "url": "https://api.pulsetic.com/mcp",
      "headers": {
        "Authorization": "YOUR_API_KEY"
      }
    }
  }
}

Cursor

Open Settings → MCP Servers → Add Server and configure:

• Type: streamableHttp
• URL: https://api.pulsetic.com/mcp
• Headers: Authorization: YOUR_API_KEY

ChatGPT

ChatGPT has native MCP support through Developer Mode. It uses OAuth to authenticate with Pulsetic.

Step 1. Enable Developer Mode in ChatGPT: go to Enabled apps → Advanced settings → Developer mode and toggle it on. This is available for ChatGPT Pro, Plus, Team, Business, Enterprise, and Edu users.

Step 2. Create a new app: go to Enabled apps → Create app.

Step 3. Fill in the app details:

Step 4. The OAuth endpoints will be auto-discovered from the server. You can verify them under Advanced settings:

Step 5. Check "I understand and want to continue" and click Create.

Step 6. Complete the OAuth login flow. You will be redirected to a Pulsetic authorization page where you enter your account email and API token.

Email must match your API token. On the authorization page, the email you enter must match the owner of the API token you provide.

Keep your API key private. Your API key has full access to your Pulsetic account. Never commit it to version control or share it publicly.

Usage Examples

Once configured, you can interact with Pulsetic through natural language.

Check on your monitors

"Show me all my monitors and their current status."

"What's the uptime for my production API monitor over the last 30 days?"

"Are any of my monitors currently down?"

"Show me the downtime events for monitor 42 in the past week."

Manage monitors

"Create a new monitor for https://api.example.com that checks every 30 seconds."

"Pause the staging monitor while we do the migration."

"Resume all paused monitors."

"Change the check frequency of monitor 15 to 60 seconds."

"Delete the old marketing-site monitor."

Handle incidents

"Create an incident on our main status page: 'Elevated API latency'
 with status investigating and message 'We're looking into slow response times.'"

"Update the API latency incident to identified: 'Root cause is a database
 connection pool issue. Working on a fix.'"

"Resolve incident 8 with message 'Connection pool limits increased.
 Response times back to normal.'"

Schedule maintenance

"Schedule a maintenance window on status page 1 called 'Database upgrade'
 starting 2025-03-20 at 2am UTC and ending at 4am UTC."

"Delete the maintenance window we scheduled for next week."

Troubleshooting

401 Unauthorized

• Verify your API key is correct and hasn't expired.
• Make sure you're passing the key in the Authorization header (either as a raw token or with Bearer prefix).
• Confirm your account is on a plan with API access enabled.

OAuth authorization fails (ChatGPT)

• On the authorization page, the email you enter must match the owner of the API token.
• Make sure Developer Mode is enabled in your ChatGPT settings.
• Ensure you have a ChatGPT plan that supports Developer Mode (Pro, Plus, Team, Business, Enterprise, or Edu).

"Monitor not found or access denied"

• The monitor ID doesn't exist or belongs to a different account.
• Use list_monitors to see your available monitor IDs.

Tools not showing in your AI client

• Restart the AI client after editing the MCP configuration.
• Check that the URL is exactly https://api.pulsetic.com/mcp.
• Test the endpoint with cURL to verify connectivity.

Getting protocol errors

• Ensure requests use Content-Type: application/json.
• The jsonrpc field must be exactly "2.0".
• Include an id field (integer or string) for all requests except notifications.