This page walks you through connecting your AI assistant to AdAdvisor’s MCP server. Pick your client below and follow the instructions.
MCP Server URL
All clients connect to the same server:
https://d3ih39u72cibno.cloudfront.net/mcp
Client setup
Claude Code
Claude Desktop
ChatGPT
Cursor
VS Code
Gemini CLI
Codex
Windsurf
Claude Code runs in the terminal and supports both OAuth and API key authentication.Option 1: OAuth (recommended)Run this command to add the server:claude mcp add adadvisor \
--transport http \
https://d3ih39u72cibno.cloudfront.net/mcp
claude mcp add adadvisor \
--transport http \
https://d3ih39u72cibno.cloudfront.net/mcp \
-- --header "Authorization: Bearer YOUR_API_KEY"
.mcp.json file in your project root:{
"mcpServers": {
"adadvisor": {
"type": "http",
"url": "https://d3ih39u72cibno.cloudfront.net/mcp",
"headers": {
"Authorization": "Bearer YOUR_API_KEY"
}
}
}
}
Claude Desktop supports MCP servers through its Connectors UI.
- Open Settings > Connectors
- Click Add Connector
- Enter the server URL:
https://d3ih39u72cibno.cloudfront.net/mcp
- Claude Desktop will open your browser to log in to AdAdvisor via OAuth
That’s it. No API key needed.Connectors are available on Claude Pro, Max, Team, and Enterprise plans.
ChatGPT supports MCP servers through its Apps settings (previously called Connectors).
- Go to Settings > Apps & Connectors > Advanced and turn on Developer Mode
- Go back to Settings > Apps & Connectors and click Create
- Enter a name (e.g., “AdAdvisor”) and the server URL:
https://d3ih39u72cibno.cloudfront.net/mcp
- ChatGPT will handle the OAuth login flow through your browser
Available on ChatGPT Pro, Team, Enterprise, and Edu plans.
Add the following to your .cursor/mcp.json file in your project root (or ~/.cursor/mcp.json for global access):With OAuth:{
"mcpServers": {
"adadvisor": {
"type": "streamableHttp",
"url": "https://d3ih39u72cibno.cloudfront.net/mcp"
}
}
}
{
"mcpServers": {
"adadvisor": {
"type": "streamableHttp",
"url": "https://d3ih39u72cibno.cloudfront.net/mcp",
"headers": {
"Authorization": "Bearer YOUR_API_KEY"
}
}
}
}
Add the following to your .vscode/mcp.json file:With OAuth:{
"servers": {
"adadvisor": {
"type": "http",
"url": "https://d3ih39u72cibno.cloudfront.net/mcp"
}
}
}
{
"inputs": [
{
"type": "promptString",
"id": "adadvisor-key",
"description": "AdAdvisor API Key",
"password": true
}
],
"servers": {
"adadvisor": {
"type": "http",
"url": "https://d3ih39u72cibno.cloudfront.net/mcp",
"headers": {
"Authorization": "Bearer ${input:adadvisor-key}"
}
}
}
}
Requires VS Code 1.101 or later with GitHub Copilot.
Add the following to your ~/.gemini/settings.json file (or .gemini/settings.json in your project root):With OAuth:{
"mcpServers": {
"adadvisor": {
"httpUrl": "https://d3ih39u72cibno.cloudfront.net/mcp"
}
}
}
{
"mcpServers": {
"adadvisor": {
"httpUrl": "https://d3ih39u72cibno.cloudfront.net/mcp",
"headers": {
"Authorization": "Bearer YOUR_API_KEY"
}
}
}
}
OpenAI’s Codex CLI uses a TOML config file.With OAuth:Add to your ~/.codex/config.toml:[mcp_servers.adadvisor]
url = "https://d3ih39u72cibno.cloudfront.net/mcp"
codex mcp login adadvisor to complete the browser login.With API Key:[mcp_servers.adadvisor]
url = "https://d3ih39u72cibno.cloudfront.net/mcp"
bearer_token_env_var = "ADADVISOR_API_KEY"
export ADADVISOR_API_KEY="your-api-key-here"
Add the following to your Windsurf MCP config file:
- macOS/Linux:
~/.codeium/windsurf/mcp_config.json
- Windows:
%USERPROFILE%\.codeium\windsurf\mcp_config.json
With OAuth:{
"mcpServers": {
"adadvisor": {
"serverUrl": "https://d3ih39u72cibno.cloudfront.net/mcp"
}
}
}
{
"mcpServers": {
"adadvisor": {
"serverUrl": "https://d3ih39u72cibno.cloudfront.net/mcp",
"headers": {
"Authorization": "Bearer ${env:ADADVISOR_API_KEY}"
}
}
}
}
Verifying the connection
After adding the server, try asking your AI assistant something about your ad data. For example:
What campaigns are currently running in my ad account?
If you get an authentication error, double-check that your API key is correct and hasn’t been revoked. You can verify active keys in Settings > MCP Server in AdAdvisor.
Finding the setup instructions in AdAdvisor
You can also find these instructions directly in the app. Go to Settings > MCP Server and scroll down to the “Quick Start” section. Select your client from the dropdown and the instructions (with code blocks you can copy) will appear.
