Access, auth, and limits
Connection is a URL and a sign-in: point your client at https://app.airops.com/mcp over streamable HTTP and authenticate with OAuth 2.0 in the browser. No key ever lands in a config file. The vendor documents the exact Claude Code command (claude mcp add --transport http airops https://app.airops.com/mcp) plus setups for Cursor, VS Code and Windsurf, and for clients that still need SSE, mcp-remote as a proxy. Probed on 2026-07-30, an unauthenticated call returns 401, so the endpoint is real and the gate works.
Access depends on your workspace rather than a published tier list: the docs say any AirOps plan with AEO enabled and at least one Brand Kit configured, and Claude's side needs a Pro, Team or Enterprise account because the free plan does not support connectors. Since AirOps publishes no prices at all, the entry cost of that arrangement is a sales conversation. Worth asking about explicitly, because the pricing table lists API data export only against Enterprise even though the MCP surface itself is not described that way.
Where it works and what to build
Claude Web and Claude Desktop get the best experience through the AirOps Connector, which adds Interactive Apps: your visibility trend arrives as a rendered chart in the conversation. Claude Code, Cursor, VS Code and Windsurf connect to the same server over plain MCP with the same tools, minus the interactive rendering. One brain, two ways in, as the vendor's own FAQ puts it.
The workflow this earns its place in is the Monday morning triage. Ask which pages are losing AI visibility, get list_pages filtered on citation_rate_decline, pull the prompts and citations behind the drop with get_prompt_answers and list_aeo_citations, then have the agent write the shortlist into a Grid with write_grid so the content run has somewhere to start. Because Brand Kit writes are exposed too, an agent can also keep tone, audiences and competitor lists current instead of leaving that to whoever remembers.
Which AI clients can use AirOps MCP?
Not every AI assistant supports MCP natively. Per-client picture for AirOps:
| Client | Support | Notes |
|---|---|---|
| Claude (Desktop/Web) | Via connectors | Official AirOps Claude Connector in Claude's directory, with Interactive Apps so charts render in the conversation. |
| Claude Code | Native MCP | Vendor documents: claude mcp add --transport http airops https://app.airops.com/mcp, then /mcp to authenticate. |
| Cursor | Native MCP | Settings, Tools and MCP, New MCP server, then Connect to complete OAuth. |
| Windsurf | Native MCP | Settings, Cascade, MCP Servers, Add custom server with the app.airops.com/mcp URL. |
How to connect AirOps MCP to Claude Desktop
Add this block to ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows), then restart Claude Desktop. Cursor and Codex use a similar schema.
{
"mcpServers": {
"airops": {
"command": "npx",
"args": ["mcp-remote", "<YOUR_MCP_URL>"]
}
}
}
// HTTP-transport MCP. Get your <YOUR_MCP_URL> from the vendor dashboard.
// Server source: https://app.airops.com/mcp
// Claude Code (one-liner): claude mcp add airops <YOUR_MCP_URL> --transport httpSome MCP servers require an API key. Check the source page for setup: app.airops.com/mcp.
Does AirOps MCP cost extra?
The AirOps MCP server is community-built — free to run, but you still pay for your AirOps subscription.
- Community MCPs are normally self-hosted from a GitHub repo. There's no extra license fee.
- Your AirOps account still needs whatever plan is required to access the data the MCP server queries.
- Trade-off: zero MCP cost, but no guarantee it'll keep working when AirOps changes their UI or API.
AirOps also publishes a REST API
AirOps publishes an API, but per-tier gating wasn't extracted automatically.
What you can do with AirOps from an AI agent
Three shapes of agent workflow. Pick the row that matches how you actually use Claude / Cursor / Codex day to day.
Skip the dashboard, ask Claude
- "What keywords moved this week?" answered in seconds
- Quick one-off questions without leaving the editor
- Pipe AirOps data into your personal notes / docs without writing a script
Run client analysis at agent speed
- Hand the MCP server to your team — every consultant can ask Claude about any client's AirOps account
- Multi-account workflows without context-switching the AirOps UI
- Reports you'd normally spend hours on, done before standup
Production agent workflows
- Embed in your internal Codex / Cursor agents that triage AirOps data daily
- Combine with other MCP servers in a multi-step agent chain
- Constrain agents to read-only roles to keep production AirOps safe
Limits and gotchas
- Official MCP server — published by AirOps themselves, so it should track product changes.
- Fallback: the REST API works for any flow the MCP server doesn't cover.
- Rate limits: MCP calls count against your underlying AirOps API quota. Burst usage from a curious agent can drain a daily allowance fast.
Agent-readiness verdict
One of the strongest MCP implementations in AI search, and a 9 because it goes past reading. An agent can pull your citation and share-of-voice data, find the pages losing visibility, then update the Brand Kit and build a Grid to act on what it found, all over one OAuth connection. The reasons it is not a 10 are commercial rather than technical: AirOps publishes no prices, and the plan comparison table lists API data export against Enterprise, so confirm what your tier exposes before you design a workflow around it.
Scored by Joonas (TMB).
AirOps MCP & API FAQ
Is the AirOps MCP server official?+
Can I use the AirOps MCP with Cursor or Codex?+
Does AirOps support webhooks?+
Sources
- AirOps official site: https://www.airops.com
- MCP server page: https://app.airops.com/mcp
Data verified by Joonas on the dates shown. MCP server status auto-rechecked weekly.
10+ years in digital marketing. I review marketing software for AI-stack fit: real pricing, MCP and API support, and how cleanly each tool drops into an AI agent workflow, cross-checked against verified data and real user feedback.