Developer tools can use Bannerproof in two ways: with your API token (simplest, works everywhere) or by signing in when the client asks (OAuth, no token to manage). Both give the same access.
Server address: https://api.bannerproof.app/mcp
1. Get your API token
Open Settings in Bannerproof and click Generate token under API token. The token is shown once — copy it. The AI assistants (MCP) block below it shows ready-to-copy commands with your token filled in. Treat the token like a password; you can regenerate or revoke it on the same page.
2. Add the server
Claude Code
claude mcp add --transport http bannerproof https://api.bannerproof.app/mcp --header "Authorization: Bearer YOUR_API_TOKEN"
Prefer signing in? Leave out the --header part, then run /mcp inside Claude Code and choose Authenticate for bannerproof.
Cursor
Add this to ~/.cursor/mcp.json (all projects) or .cursor/mcp.json (one project):
{
"mcpServers": {
"bannerproof": {
"url": "https://api.bannerproof.app/mcp",
"headers": { "Authorization": "Bearer YOUR_API_TOKEN" }
}
}
}
VS Code
Add this to .vscode/mcp.json in your workspace (or your user MCP configuration) and start the server from the MCP view. VS Code offers to sign in when it starts the server.
{
"servers": {
"bannerproof": { "type": "http", "url": "https://api.bannerproof.app/mcp" }
}
}
Other clients
Any client that supports remote MCP servers over HTTP works: use the server address above and send the header Authorization: Bearer YOUR_API_TOKEN, or let the client run the OAuth sign-in.
3. Use it while you code
The real advantage in an editor: the assistant can see your code. Ask it to "check staging.example.com, then fix what fires before consent in our GTM setup / consent config / theme", and to re-check afterwards with scan_domain. See What you can ask for all tools.
Keep the token out of shared repositories — put project-level config with a token in .gitignore, or use the sign-in instead.