openstatus logoPricingDashboard

Connect Openstatus to Claude Desktop (MCP Setup Guide)

May 08, 2026 | by openstatus | [education]

TL;DR — Add https://api.openstatus.dev/mcp as a custom connector, approve it in the browser, and Claude Desktop can drive your status pages: list incidents, draft updates, and schedule maintenance from the chat window. No config file, no key to paste. Every mutation is audit-logged under actor_type = 'mcp' and notifications never fire unless you explicitly approve.


Claude Desktop speaks the Model Context Protocol (MCP) — the open standard for letting AI assistants reach external services with explicit approval on every call. This guide wires the openstatus MCP server into Claude Desktop so you can list status pages, post incident updates, resolve reports, and schedule maintenance windows without leaving the chat window. The same steps work in Claude.ai in the browser: the two share one connector list.

Why connect Openstatus to Claude Desktop?

If your incident workflow already runs through chat, Claude Desktop is the right drafting surface. Concrete scenarios:

  • Drafting customer-facing copy — turn engineer shorthand ("payment-api spiking p99 to 6s") into the calm, structured update your subscribers expect, then approve and post it without leaving the chat.
  • On-call summaries — "Summarize the last 7 days of incidents on my API page" in seconds, useful for weekly reviews and handoffs.
  • Read-only audits — a teammate connects with read-only access; their Claude Desktop sees only the list_* and get_* tools and cannot mutate anything.

The same server works with Claude Code, ChatGPT, Codex, opencode, Cursor, and any other MCP-compatible client.

Prerequisites

Step 1: Add a custom connector

In Claude Desktop, open Settings > Connectors and click Add custom connector. Fill in:

  • Name: openstatus
  • Remote MCP server URL: https://api.openstatus.dev/mcp

Leave the advanced OAuth fields empty. openstatus registers the connector automatically; there is no client id or secret to enter.

Step 2: Approve the connection

Click Connect. Your browser opens the openstatus consent screen:

  • Workspace — the workspace this connector can reach. If you belong to several, pick one; each workspace is a separate connection.
  • AccessRead-only lets Claude list status pages, reports, monitors, and maintenance windows. Read & write also lets it create reports, append updates, resolve incidents, and schedule maintenance.

Click Approve. You land back in Claude Desktop with the connector marked as connected. Tokens refresh in the background; you won't be asked again unless the connection is revoked.

Step 3: Enable the tools in a chat

Open a new chat and click the tool picker. Switch the openstatus connector on. With read-only access you'll see list_status_pages, list_status_reports, list_maintenances. With read & write you'll also see create_status_report, add_status_report_update, update_status_report, resolve_status_report, and create_maintenance.

Step 4: Run your first prompt

Try a read tool first:

List my openstatus status pages

Claude calls list_status_pages, asks for approval, and returns the slugs and ids in your workspace. With write access, draft an incident — Claude shows you the title, status, message, and notify choice before firing:

Draft a status report on my "api" page: investigating elevated latency on the payment endpoint, do not notify subscribers yet

Alternative: API key with mcp-remote

If you'd rather authenticate with an API key — a shared machine, or a key that outlives your user — Claude Desktop can bridge the server through mcp-remote. You'll need Node.js 18+ on your PATH. Create a key in Settings > API Tokens (scopes are immutable on a key), then add the server to claude_desktop_config.json (Settings > Developer > Edit Config):

{
  "mcpServers": {
    "openstatus": {
      "command": "npx",
      "args": [
        "mcp-remote",
        "https://api.openstatus.dev/mcp",
        "--header",
        "x-openstatus-key: ${API_KEY}"
      ],
      "env": {
        "API_KEY": "os_xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx"
      }
    }
  }
}

The key lives in env, not inlined in args, because mcp-remote parses --header arguments before shell expansion. Quit Claude Desktop fully (Cmd+Q on macOS, File > Quit on Windows) and relaunch; MCP servers from the config file only load at launch.

Built-in safeguards

Four properties make this safe enough for a real incident:

  • Workspace-scoped consent — the token only reaches the workspace you approved, with the access level you chose. A connector that asks for write access can be downgraded to read-only on the consent screen.
  • Required notify: boolean on every mutation tool — the schema rejects calls that omit it, so the LLM has to ask before firing. notify: false is a hard guarantee that no subscriber alert goes out.
  • Lookup-before-mutatecreate_status_report and create_maintenance require a pageId resolved via list_status_pages rather than guessed.
  • Audit log attribution — every mutation lands in the audit log under actor_type = 'mcp', with actor_id (the connection) and actor_user_id (the user who approved it), so AI-driven actions are traceable as a separate slice. Connecting and revoking are audited too.

Disconnecting

Open Settings > Integrations > Connected apps in the openstatus dashboard. Each connection shows the client, its access, who connected it, and when it was last used. Revoke it and the connector's next call fails; reconnect from Claude Desktop's Settings > Connectors. Members revoke their own connections; owners and admins can revoke any. Removing a member from the workspace revokes their connections automatically.

Troubleshooting

  • The connector shows as disconnected. Open Settings > Connectors, click openstatus, and reconnect. This is expected after a revoke from the dashboard.
  • Tools don't show up in a chat. Connectors are enabled per chat: open the tool picker and switch openstatus on.
  • "Tool not available" when creating an incident. Your connection is read-only. Reconnect and pick Read & write on the consent screen.
  • mcp-remote tools don't appear after restart. Validate the JSON (a trailing comma silently disables the whole mcpServers block) and confirm npx is on the PATH that Claude Desktop sees — nvm-managed Node sometimes isn't. Check Settings > Developer > Open Logs Folder for mcp-remote stderr.

What's next?

Frequently asked questions

Does Claude Desktop support remote MCP servers like openstatus?

Yes. Claude Desktop and Claude.ai share a connector list. Add openstatus under Settings > Connectors > Add custom connector with the URL https://api.openstatus.dev/mcp; the connection is authorized with OAuth in your browser. No local process, no Node.js, no config file.

Do I need an API key to connect Claude Desktop?

No. The openstatus MCP server is an OAuth 2.1 authorization server. Claude Desktop opens a consent screen where you pick a workspace and an access level; tokens are issued and refreshed automatically. If you prefer an API key, you can still bridge the server through npx mcp-remote with an x-openstatus-key header in claude_desktop_config.json.

Where do I revoke Claude Desktop's access?

In the openstatus dashboard under Settings > Integrations > Connected apps. Revoking signs the connector out immediately; reconnect from Claude Desktop's connector settings. Members revoke their own connections, owners and admins can revoke any.

Why don't I see the openstatus tools in a chat?

Connectors are enabled per chat. Open the tool picker in the chat and switch the openstatus connector on. If the connector shows as disconnected in Settings > Connectors, click it and reconnect.

Can Claude Desktop accidentally notify subscribers when posting an incident?

No. Every publishing tool (create_status_report, add_status_report_update, resolve_status_report, create_maintenance) requires an explicit notify: true | false field. Claude must show the notify choice before firing the tool, so an LLM cannot quietly fan out an alert by omitting the flag. update_status_report is metadata-only and has no notify path at all.

How do I give Claude Desktop read-only access to my workspace?

Pick Read-only on the consent screen. The MCP server filters write tools out of the tools/list response for read-only connections, so Claude Desktop sees only the 14 read tools — status pages and page components, status reports, maintenances, monitors, response logs, notifications, private locations, and audit logs. None of the create_*, update_*, or resolve_* tools are registered for that session.

Where do MCP-driven changes appear in the audit log?

Every mutation routed through Claude Desktop lands in the audit log under actor_type = 'mcp', with actor_id set to the connection (oat_<id> for OAuth, the key id for an API key) and actor_user_id set to the user who approved the connection or created the key. This separates AI-driven actions from CLI, dashboard, and direct API mutations.


Start free. No credit card required. Configure Claude Desktop to drive your status pages in under 5 minutes.

Try openstatus free