How to use Arenza MCP with Claude Code and Cursor

Model Context Protocol (MCP) is Anthropic's open standard for connecting language models to external tools. Arenza packages its intent-to-growth workflow as a marketer-facing MCP: the marketer speaks naturally, while Claude calls structured Arenza capabilities underneath.

What the Arenza MCP server can do

list_brand_contexts({}) — list the authorized brands you can continue working on; it does not score or scan them.
understand_brand({website}) — remember the brand, offerings and verified official-site facts; no measurement is started.
map_buyer_intents({brand_context_id, questions?}) — map real buyer questions to marketer-friendly roles, intents, funnel stages and channels.
measure_intent_performance({brand_context_id, intent_ids?, freshness?}) — omit freshness or use latest_available to read stored AVS/ACS without refreshing. Use current only when the marketer explicitly asks to refresh or re-measure; Arenza reuses a current result, completes stored-answer ACS, or starts one eligible fresh measurement.
find_growth_opportunities({brand_context_id}) — rank measured intents by visible growth factors and distinguish measured revenue from directional demand.
prepare_growth_content({opportunity_id, formats}) — prepare grounded Buyer Guides, product-page proposals, YouTube scripts or Reddit posts; channel drafts include a publication target and nothing publishes automatically.

Advanced correction, article-scoring and single-page audit tools remain available for technical integrations. Marketer workflow tools are scope-gated: brand setup, intent changes, first/current measurement and content drafts appear only after the user approves those permissions. Generic status questions remain reads; only an explicit refresh request may use measurement allowance.

Setup for Claude Desktop

Open Claude Desktop → Settings → Developer → Edit Config.
Add the Arenza MCP server entry to mcpServers (see config below).
Restart Claude Desktop.
In a new chat, type "list my brands" — Claude will call list_brand_contexts and render the selector.

{
  "mcpServers": {
    "arenza": {
      "url": "https://api.arenza.ai/mcp"
    }
  }
}

That's the whole entry — just the URL. On the first call Claude opens a browser to sign in and approve access; it auto-discovers the OAuth endpoints (RFC 8414 + 9728), so there's no token to paste and no header to manage.

Setup for Claude Code (CLI)

claude mcp add --transport http --scope user arenza https://api.arenza.ai/mcp
claude mcp login arenza
claude mcp list

Setup for Cursor

Add the same URL to .cursor/mcp.json, then approve and authenticate the server:

{
  "mcpServers": {
    "arenza": {
      "url": "https://api.arenza.ai/mcp"
    }
  }
}

agent mcp enable arenza
agent mcp login arenza
agent mcp list-tools arenza

Sample queries Claude can answer once connected

"Here is our website. Which buyer intents could grow revenue for us?"
"Map these buyer questions to the roles that would ask them."
"Measure AVS and ACS for our steep-slope mower intents."
"Prepare a YouTube script and Reddit post for the strongest opportunity, and tell me where to publish each."

Why MCP-native matters for agencies

Most GEO tools ship a dashboard-only experience. That's fine if your team has 1 brand to look after. If you have 20, you don't want to context-switch between tabs — you want to ask Claude "summarize all client brands' AI visibility for the Monday standup" and have it done in one prompt. MCP makes that possible.

Safety boundary

Asking to measure may start only the included first measurement; existing evidence never triggers an extra or paid rescan.
AVS and ACS always include coverage, sample and freshness; no valid observation is measuring/unavailable, not zero.
Content generation produces review-only artifacts. Publish and product-page writes remain behind separate human approval and live verification.