One workspace endpoint
A single MCP endpoint covers your workspace. Each project-level tool accepts a project_id argument to scope the call.
Features · MCP Server
Connect AI agents to your Ceyo data .
Connect MCP-compatible AI tools to Ceyo visibility data, prompts, actions, competitors, and analytics.
MCP endpoint Ceyo MCP
Same API key
POST /mcp MCP
JSON-RPC 2.0
What MCP unlocks
Give agents the context to act.
Ceyo MCP lets AI tools read and work with the same project state your team uses in the platform, scoped by workspace API key and project ID.
Same API key
The MCP server uses the same Ceyo API key as the REST API, with X-Api-Key or Authorization bearer headers.
Full visibility context
Expose visibility metrics, prompts, LLM responses, citations, competitors, actions, fan-out queries, and analytics.
Agent-ready workflows
Connect Cursor, Claude Desktop, n8n, or a generic MCP client so agents can inspect Ceyo data directly.
01 · Clients
Connect the AI tools your team already uses.
The MCP docs include setup paths for Cursor, Claude Desktop, n8n, and generic MCP clients, with OAuth-based remote clients marked as coming soon.
- Cursor deeplink. Install the Ceyo MCP server entry in Cursor, then replace the API key placeholder with a workspace key.
- Claude Desktop. Manual local config is live today through mcp-remote and the Ceyo MCP endpoint.
- n8n and generic clients. Use the MCP Client node or copy the endpoint and X-Api-Key header into any MCP-compatible client.
02 · Tool coverage
Expose the project state agents need for GEO work.
The MCP server is not just a metrics endpoint. It gives agents structured access to workspace, project, prompt, action, competitor, suggestion, fan-out, and analytics data.
- Project and prompt intelligence. Read project metrics, visibility time series, citations, prompt responses, prompt citations, and prompt competitors.
- Optimization actions. List actions, inspect action detail, and update action status where write access is available.
- Analytics context. Use GA4-backed analytics tools for overview, conversions, landing pages, referrals, and conversion events when connected.
03 · Protocol and security
Built on MCP and JSON-RPC with predictable scoping.
The server implements Model Context Protocol over HTTP using JSON-RPC 2.0. Authentication happens before tool calls, and each project operation is scoped by project_id.
- MCP protocol. Supports initialize, tools/list, tools/call, ping, and notification messages over one HTTP endpoint.
- Workspace authentication. Use X-Api-Key or Authorization bearer headers with your Ceyo workspace API key.
- Clear error layers. HTTP errors, JSON-RPC errors, and tool errors each have a distinct shape for agents to handle.
Setup flow
From API key to agent context.
Create an API key
Create or reuse a workspace API key from Workspace settings, the same credentials used by the REST API.
Add the endpoint
Configure your MCP client with the Ceyo MCP endpoint and pass the API key through an accepted header.
List projects
Ask the agent to list projects to confirm the workspace connection and available project IDs.
Call tools
Use project_id scoped tool calls to inspect visibility, prompts, actions, competitors, fan-out queries, and analytics.
MCP Server FAQ.
It exposes Ceyo workspace and project tools covering visibility metrics, prompt tracking, LLM response intelligence, citations, competitors, actions, fan-out queries, suggestions, and GA4 analytics when connected.
The docs include setup paths for Cursor, Claude Desktop, n8n, and generic MCP clients. ChatGPT and Claude.ai remote setup require OAuth support and are marked as coming soon.
No. The MCP server uses the same Ceyo workspace API key as the REST API, passed through X-Api-Key or Authorization bearer headers.
The endpoint is workspace-scoped by API key. Project-level tools accept a project_id argument so the same MCP connection can work across multiple projects.
Most tools are read tools, but selected write tools exist, such as creating projects, creating prompts, and updating action status, depending on access and plan permissions.
Agent-ready data access
Bring Ceyo data into your AI tools.
See how the MCP server lets your agents inspect visibility, prompts, competitors, actions, and analytics without switching back to the dashboard.