CitationBench is a hosted MCP server for marketing teams and developers, exposing ~35 battle-tested SEO and GEO tools so you can build powerful agents without months of API integration.
# Add CitationBench hosted MCP server to Claude Code> claude mcp add citationbench --url https://mcp.citationbench.com/mcp✓ Connected · 35 tools available# Research keywords directly from your agentresearch.keyword.research> { "seed": "marketing automation tools", "limit": 20, "label": true }{ "keywords": [{ "kw": "ai marketing agent", "intent": "PRODUCT", "relevance": "CORE" }, ...# Bootstrap a new client brand in one agent callagent.invoke> { "agent": "brand-bootstrap", "input": { "url": "https://newclient.com" } }job_id: job_4mz9q · status: running · awaiting_approvalStitching together raw APIs for every SEO task wastes engineering time and leaves agents fragile at scale.
Building on raw APIs from Ahrefs, Surfer, and Apollo separately means months of glue code before your agent does anything useful.
Generic API providers give you endpoints but no composable SEO workflow — research, produce, publish, and index are all disconnected.
Running your own MCP infrastructure means every restart, patch, or scaling event falls on your team to handle.
The fix
CitationBench's hosted MCP server at mcp.citationbench.com/mcp gives developers and agencies ~35 production-ready tools across research, content production, distribution, and agent orchestration — callable from Claude Code, Cursor, Windsurf, and ChatGPT Apps today.
Five feature areas cover the full marketing ops workflow — from keyword discovery and content production to indexing, link building, and durable agent jobs.
MCP Infrastructure
Connect to mcp.citationbench.com/mcp in one command. Tools span research.*, produce.*, indexing.*, link_building.*, and agent.* namespaces — no infrastructure to manage.
Agent Orchestration
Call agent.invoke from any MCP client to run named agents like brand-bootstrap. Supports multi-turn continuation, real-time SSE streaming, and human approval gates mid-workflow.
Research Tools
Run research.keyword.research directly from your MCP client. Discovers, clusters, and labels keyword sets from a seed URL. Results persist automatically for content production steps.
Agency Architecture
Manage every client brand from one master API key. Pass X-Workspace-Id to isolate keywords, content, campaigns, and settings per client — no separate credentials per workspace.
Reliability
All agent invocations are backed by durable Cyclonic workers. Cancel via agent.invocations.cancel, resume via agent.invocations.continue. Jobs survive process restarts and are fully auditable.
How it works
An `sk_test_*` key lands in your dashboard instantly. No demo gate — start calling real endpoints with shape-complete responses.
`claude mcp add citationbench https://mcp.citationbench.com/mcp` from Claude Code, Cursor, or any MCP client. Or `curl` against `api.citationbench.com/v1/*` directly.
Every tool returns shape-complete demo data without auth, so your agent works before the user signs up. Tools across research, production, indexing, link-building, and agents.
Add `X-Workspace-Id: ws_***` to scope per client. Switch to a live `sk_live_*` key when you're ready. Same API surface, same SDK.
Why CitationBench
Hosted at mcp.citationbench.com/mcp. Works with Claude Code, Cursor, Claude Desktop, Windsurf, and ChatGPT Apps. No self-hosting required.
Every long-running call runs on Cyclonic workers — survives restarts, cancellable, resumable, streams via SSE. Production-safe by default.
One master API key, N client workspaces. Switch with a single `X-Workspace-Id` header. Per-client data isolation, bulk ops across all of them.
Every endpoint responds in shape-complete demo mode without auth, so you can build the agent before the user signs up.
FAQ
Connect the hosted MCP server for marketing to your agent in minutes and access ~35 SEO tools without self-hosting.
View MCP Docs