Connect AI Assistants to AEO Optima

Why Connect AI Assistants?

AEO Optima exposes a Model Context Protocol (MCP) server with 112 tools that let AI assistants directly access your brand visibility data, capture snapshots, run analyses, manage goals, and generate reports — all from within your AI tool of choice.

Instead of switching between your AI assistant and the AEO Optima dashboard, you can ask questions like:

• "How is my brand visibility trending this month?"
• "Which competitors are gaining on us?"
• "Run a comprehensive AI analysis on my latest snapshot data"
• "Create a goal to reach 60% visibility by Q3"
• "Generate a PDF report for the leadership team"

Your AI assistant handles the query using AEO Optima's data and intelligence engines.

Supported AI Assistants

Quick Setup (Any Client)

Step 1: Generate a Token

1. Log in to AEO Optima
2. Go to Settings → MCP / API
3. Click Generate New Token
4. Choose a role cap (viewer, member, or admin)
5. Copy the token — it starts with aeo_ and is shown only once

Step 2: Configure Your AI Client

Server URL: https://aeo-optima-mcp.onrender.com/mcp

Transport: Streamable HTTP (JSON-RPC over HTTP POST)

Add these to your AI client's MCP configuration:

{
  "mcpServers": {
    "aeo-optima": {
      "url": "https://aeo-optima-mcp.onrender.com/mcp",
      "transport": "streamable-http",
      "headers": {
        "Authorization": "Bearer aeo_your_token_here"
      }
    }
  }
}

Step 3: Start Using It

Once connected, your AI assistant has access to 112 tools across these categories:

Client-Specific Setup Guides

Claude.ai (web, mobile, desktop) — custom connector

Claude.ai connects with OAuth — no token to paste, and every AEO Optima role and plan gate applies to the connection automatically.

1. In Claude.ai open Settings → Connectors → Add custom connector
2. Name it (for example AEO Optima) and paste the URL: https://aeo-optima-mcp.onrender.com/mcp
3. Leave "OAuth Client ID" and "Client Secret" empty — AEO Optima registers the connector for you (dynamic client registration)
4. Click Add, then Connect. An AEO Optima sign-in page opens: sign in, pick the organisation and access level, click Authorize
5. Back in Claude.ai the connector shows a green check. Try: "List my AEO Optima projects."

If Claude.ai says "Your connection … stopped working. Reconnect to continue" — the access grant expired or was revoked (Settings → MCP → Revoke does this on purpose). Click Reconnect and authorize again; nothing needs re-adding.

If it says "Couldn't register with … sign-in service" — Claude.ai could not reach the AEO Optima sign-in service. This is never fixed by entering a Client ID; try again in a minute, and if it persists contact support with the reference shown.

Claude Desktop

The simplest setup — supports OAuth for click-to-connect:

1. Open Claude Desktop → Settings → MCP Servers
2. Click Add Server
3. Enter the URL: https://aeo-optima-mcp.onrender.com/mcp
4. Choose OAuth (recommended) or paste your Bearer token
5. Claude will show available tools — you're connected

Claude Code (CLI)

claude mcp add aeo-optima \
  --transport http \
  --url https://aeo-optima-mcp.onrender.com/mcp \
  --header "Authorization: Bearer aeo_your_token_here"

Cursor

Add to .cursor/mcp.json in your project root:

{
  "mcpServers": {
    "aeo-optima": {
      "url": "https://aeo-optima-mcp.onrender.com/mcp",
      "transport": "streamable-http",
      "headers": {
        "Authorization": "Bearer aeo_your_token_here"
      }
    }
  }
}

ChatGPT

1. Go to Settings → Developer Mode → MCP Servers
2. Add a new server with the URL and Bearer token
3. ChatGPT will discover all 112 tools automatically

OpenAI Codex

One config in ~/.codex/config.toml covers the Codex CLI, the Codex IDE extension, and ChatGPT desktop. OAuth is the recommended path (codex mcp login aeo-optima opens the AEO Optima consent screen), with a static-token fallback for headless machines. See the MCP API Reference → OpenAI Codex for both config snippets.

There is also a public plugin that installs the connection plus an AEO workflow skill in one step — the same marketplace serves both OpenAI Codex and Claude Code:

codex plugin marketplace add teachskillofskills-ai/aeo-optima-codex-plugin

VS Code with Copilot

Add to .vscode/mcp.json:

{
  "servers": {
    "aeo-optima": {
      "type": "http",
      "url": "https://aeo-optima-mcp.onrender.com/mcp",
      "headers": {
        "Authorization": "Bearer aeo_your_token_here"
      }
    }
  }
}

What You Can Do

Monitor Mention Rate & Visibility Score

"What's my mention rate and visibility score, and how have they changed this week?"

The assistant calls get_dashboard_metrics and get_analytics to show your current Mention Rate, composite Visibility Score, sentiment, and period-over-period changes (a 30-day window by default, matching the dashboard).

Track Competitors

"How does my mention share compare to [competitor]?"

Uses get_analytics with competitor data to show relative positioning, trend direction, and market entropy.

Set and Track Goals

"Set a goal to reach 55% visibility for non-branded queries by September"

Calls create_goal with metric, segment, target, and date. Returns feasibility assessment and auto-computed milestones.

Run Deep Analysis

"Run a comprehensive AI analysis on my last 30 days of data"

Triggers run_analysis with type comprehensive. Returns sentiment drivers, content gaps, and scored opportunities.

Generate Reports

"Generate a standard monthly report as PDF"

Calls generate_report with project, type, format, and period. Returns a download URL and AI Brand Score.

Check Intelligence Scores

"What are my intelligence scores?"

Calls get_intelligence_scores to show all 6 proprietary metrics: Brand Narrative Coherence, Cross-Model Consistency, Market Entropy, Sentiment Drift, Citation Influence, and Entity Authority.

Rate Limits

Rate limits are per-token and tiered by your organization's plan — from 60 requests/minute, 500 requests/hour, and 5 snapshot captures/hour on Pro-Individual up to 600 requests/minute, 50,000 requests/hour, and 500 captures/hour on Custom. See the MCP API Reference for the full per-tier table; the aeo://rate-limits MCP resource reports every tier's caps plus your org's effective limits.

If you hit a limit, the server returns a 429-style error with a retryAfterSeconds value.

Security

• All tokens are encrypted at rest
• Tokens have role caps — a viewer token can read data but not modify anything
• OAuth 2.1 with PKCE is supported for browser-based flows
• All API calls are logged in the MCP audit log (viewable in Settings)
• HMAC-SHA256 webhook signatures for outbound events

Related

• MCP API Reference — complete list of 112 tools with parameters and roles
• Webhooks — 11 real-time event types for automation
• API Keys — manage tokens and permissions
• Plans — feature availability by plan tier