Agent setup

Connect your AI coding agent to Hyperping through the hosted MCP server. Once connected, the agent can check what is down, create and update monitors, pull SLA numbers, and investigate outages without leaving your editor or terminal.

Set up in one prompt

Copy the prompt below and paste it into Claude Code, Cursor, Windsurf, Codex, or any agent that supports MCP. It walks the agent through the whole setup, including verification.

Set up the Hyperping MCP server for me.

1. Add an MCP server named "hyperping" using streamable HTTP transport:
   URL: https://api.hyperping.io/v1/mcp
   Header: Authorization: Bearer HYPERPING_API_KEY
   Write it into this client's MCP config. If you cannot edit the config
   directly, print the exact block and the file path where it goes.
2. Ask me for my API key and replace HYPERPING_API_KEY with it. I can
   create one in Hyperping under Project Settings > API Keys.
3. Fetch https://hyperping.com/mcp/llms.txt to learn the available tools
   and query patterns.
4. Verify the setup by calling get_status_summary and report how many
   monitors are up and down.

Manual setup per agent

Prefer to edit the config yourself? Each client stores its MCP servers in a different place. The server details are the same everywhere: streamable HTTP, https://api.hyperping.io/v1/mcp, Bearer auth.

Claude Code

One command from your terminal:

claude mcp add --transport http hyperping https://api.hyperping.io/v1/mcp \
  --header "Authorization: Bearer sk_your_key"

Add --scope project to write the server into a .mcp.json checked into the repo, so your whole team gets it. Run /mcp inside a session to confirm the connection.

Cursor

Create .cursor/mcp.json at the root of your project, or ~/.cursor/mcp.json for all projects:

{
  "mcpServers": {
    "hyperping": {
      "url": "https://api.hyperping.io/v1/mcp",
      "headers": {
        "Authorization": "Bearer sk_your_key"
      }
    }
  }
}

Claude Desktop

Open Settings → Developer → Edit Config and add the server to claude_desktop_config.json:

{
  "mcpServers": {
    "hyperping": {
      "type": "http",
      "url": "https://api.hyperping.io/v1/mcp",
      "headers": {
        "Authorization": "Bearer sk_your_key"
      }
    }
  }
}

Windsurf

Add the server to ~/.codeium/windsurf/mcp_config.json:

{
  "mcpServers": {
    "hyperping": {
      "serverUrl": "https://api.hyperping.io/v1/mcp",
      "headers": {
        "Authorization": "Bearer sk_your_key"
      }
    }
  }
}

Codex CLI

Codex talks to MCP servers over stdio, so bridge to the hosted server with mcp-remote in ~/.codex/config.toml:

[mcp_servers.hyperping]
command = "npx"
args = ["-y", "mcp-remote", "https://api.hyperping.io/v1/mcp", "--header", "Authorization: Bearer sk_your_key"]

Try it

Once connected, ask in plain language. A few prompts that map directly to tools:

• "What is down right now, and since when?"
• "Create an HTTP monitor for api.example.com, checked every 30 seconds from London and Frankfurt."
• "Which monitors had the worst uptime this month? Give me the SLA numbers."
• "Show me the timeline of the last outage on the checkout monitor."
• "Who is on call right now?"

The server exposes 25 read tools and 5 write tools. The full list, with query patterns and response shapes, is on the MCP Integration page.

Give your agent standing context

Two URLs make any agent better at working with Hyperping. Reference them in your agent's project instructions (CLAUDE.md, .cursor/rules, or equivalent):

• https://hyperping.com/mcp/llms.txt describes every MCP tool, its parameters, and common query patterns.
• https://hyperping.com/docs/llms.txt indexes the full documentation, so the agent can look up any feature on demand.

Monitoring as code

For monitors that should live in version control, use the Terraform provider instead of MCP writes. Agents handle it well: describe the monitor you want and ask for the HCL, then review the plan like any other infrastructure change.

Security model

1. Keys are project-scoped

   An API key belongs to exactly one project. Cross-project access returns 404.

2. Read-only is enforced server-side

   Write tools refuse read_only keys at the tool handler. The agent cannot talk its way around it.

3. Revocation is immediate

   Deactivate a key in the dashboard and the next MCP call fails. No token lifetime to wait out.

Deletion is deliberately not exposed over MCP. Agents can pause monitors but never destroy them. Details on the MCP Integration page.

Next steps