Two clicks in your OpenGraphy settings and Claude becomes a power user of your entire marketing stack — running AI visibility audits, drafting articles, generating images, scheduling content, even publishing — without ever opening the app. The OpenGraphy MCP server ships with 50+ functions, all wired to your brand context.
The Model Context Protocol (MCP) is the open standard, championed by Anthropic, that lets AI assistants safely call external tools. In 2024 it was a niche developer protocol; by mid-2026 it has become the way serious AI assistants integrate with SaaS platforms. OpenGraphy was one of the first content-marketing platforms to ship a production MCP server, and the surface area is now broad enough that Claude can drive most of the platform end-to-end from a single conversation.
This guide walks through what the OpenGraphy MCP server actually does: the function categories, the workflow scenarios it unlocks, the activation steps (literally two clicks), and the security model. By the end you should be able to operate OpenGraphy from inside Claude — including for use cases that previously required jumping between five tabs.
What is the MCP?
The Model Context Protocol is the open standard that lets an assistant like Claude discover and call external tools safely. In practice, the assistant queries the connected MCP server to learn what functions are available, decides which one fits the user's request, calls it with the right arguments, and incorporates the result into its response.
Why MCP and not a custom integration
Before MCP, every SaaS platform that wanted AI integration had to build a custom adapter for ChatGPT, another for Claude, another for Gemini. MCP collapses that to one server that any compatible assistant can use. Same protocol; many clients. Less brittle than vendor-specific integrations.
What "safely" means here
Three layers of safety. Authentication: the MCP server requires your OpenGraphy API key, scoped to your account. Permission boundaries: every function has an explicit permission scope; the assistant can't escalate. Auditability: every call is logged in your OpenGraphy account, viewable in the integration log.
What Claude can do inside your account, today
The OpenGraphy MCP server exposes 50+ functions across four major categories. Here's the surface area, by capability.
Brand intelligence (~10 functions)
List your brands and read full Brand DNA (palette, fonts, voice, values, audience). Analyze any brand from a URL. Pull personas, topical clusters, and brand-level statistics. Useful for "Claude, summarize the voice profile of brand X" or "Claude, audit a competitor brand from this URL."
AI visibility and audits (~15 functions)
Run AI visibility audits and read details (rankings, gaps, cited sources). Inspect competitors detected by AI models (share of model, sentiment, authority). Track AI positions per query, per model, with strategic fixes. Generate stress tests, run query analyses. Useful for "Claude, audit our brand's AI visibility, identify the three biggest gaps, and propose articles to close them."
Content generation (~15 functions)
Generate articles, social posts, images, videos, music — all brand-DNA-aware. List media assets. Pull saved prompts. Generate strategy documents. Useful for "Claude, write three articles to close the gaps you identified, generate a hero image for each, and prepare social variants for LinkedIn."
Distribution and scheduling (~10 functions)
Schedule content, publish to connected platforms, manage media. Useful for "Claude, schedule the three articles across the next three weeks and publish the social variants on Tuesdays at 7:48 a.m."
The scenario that changes everything
The capability list is impressive, but the unlock is what happens when you chain functions in one prompt. Here's the scenario most B2B teams discover first, and the one that changes how they think about content operations.
One prompt, full campaign
Inside Claude, the user types: "Audit my brand's AI visibility, identify the three biggest gaps, write a 1,500-word article for each, generate a hero image, and schedule the WordPress publication across the next three weeks."
Claude chains the functions on its own: get_ai_visibility_audits → get_brand_queries → generate_article ×3 → generate_image ×3 → schedule_content ×3. Total elapsed time: about 4 minutes. The user never opened the app.
Why this matters for B2B operations
The friction in most marketing operations is not "the work itself" — it's the context-switching across tools. AI visibility data lives in one tool, content generation in another, scheduling in a third. MCP collapses the toolchain into one conversation. The work that used to take three apps and an hour now takes one prompt and four minutes.
Activate it in 2 minutes
The activation flow is shorter than reading this paragraph.
Step 1 — Get your MCP credentials
Open OpenGraphy → Settings → Integrations → Claude. The page shows your MCP server URL and API key, ready to copy.
Step 2 — Add the connector in Claude
In Claude (claude.ai web or Claude Desktop), go to Settings → Connectors → Add custom connector. Paste the URL and key. Save.
Step 3 — Confirm the connection
Ask Claude "what OpenGraphy functions can you call?" The assistant should list the function categories. You're live.
Why this is a game changer
The cinematic version: chaotic intern with tangled cables on the left, calm operator at an organized green-lit desk on the right. The reality is more boring and more important: MCP turns AI assistants from "answer questions about your work" into "do your work, with audit trails."
Three capabilities the MCP unlocks
Cross-tool orchestration — Claude can chain OpenGraphy functions with calls to your CRM, calendar or analytics platform via their respective MCP servers. Conversational operations — your team operates the platform from inside the conversation that's already open, not from yet another browser tab. Audit-trail by default — every action Claude takes is logged in OpenGraphy's integration log, so compliance and security teams can review.
What you'll use it for in week one
The most common day-1 use cases: ad-hoc AI visibility audits, drafting content for new product launches, regenerating an article that under-performed, scheduling a campaign while traveling. The friction is so low that it changes which workflows you bother to run at all.
FAQ: The OpenGraphy MCP Server
What is the Model Context Protocol (MCP)?
An open standard, championed by Anthropic, that lets AI assistants safely discover and call external tools. The OpenGraphy MCP server exposes 50+ platform functions to any MCP-compatible assistant — currently Claude (claude.ai and Claude Desktop), with broader client support arriving.
Which AI assistants can use the OpenGraphy MCP server?
Currently: Claude (claude.ai and Claude Desktop). The protocol is open, so additional clients (ChatGPT custom GPTs, Cursor, IDE assistants, etc.) work once they ship MCP-compatible connectors.
Is it secure?
Yes. Authentication uses your scoped OpenGraphy API key. Function permissions are explicit per-function. Every call is logged in your OpenGraphy integration log. Revoke access at any time from Settings → Integrations.
What does it cost?
The MCP server itself is included with your OpenGraphy plan. Function calls (image/video generation, etc.) consume the same credits they would from the OpenGraphy UI — no extra charge for the protocol layer.
Can I use it for multi-brand setups?
Yes. Each brand has its own brand_id; the MCP server propagates brand context automatically when you specify the brand in the prompt. See AI Agent Best Practices for multi-brand operating patterns.
What's the difference between this and using ChatGPT to ask about marketing?
Massive. ChatGPT alone can answer questions about marketing. Claude with the OpenGraphy MCP can do the marketing work — run audits, generate content, schedule publication — using your actual data and shipping results to your real platforms.
Going Further
- Introduction to AI Agents for Marketing — the broader agent landscape.
- Building Effective Agent Workflows — orchestration patterns.
- Understanding Your AI Visibility Score — the metric the MCP can audit on demand.
- Autonomous Content Creation with AI Agents — the full operating model.
- AI Agent Best Practices for Marketing Teams — the safe rollout path.
Ready to put this into action?
Explore our Agents features and start optimizing today
