Use your credit card intelligence anywhere — in Claude, ChatGPT, Cursor, VS Code, and more. Ask natural language questions about your cards and get instant, personalized answers.
How It Works
MCP (Model Context Protocol) lets AI assistants securely access your RewardSmart data in real time.
Generate an MCP API token in the RewardSmart app under Settings → Integrations. Requires an Elite subscription.
Follow the setup guide for your AI assistant. Paste one JSON config block and restart the app — that's it.
Ask your AI which card to use, what offers are expiring, or how to maximize rewards on your next purchase.
MCP access is available on the RewardSmart Elite plan. View pricing →
Generate your token in the RewardSmart app under Settings → Integrations → MCP Access.
You need the RewardSmart app installed with at least one card added to your wallet.
Supported Platforms
Step-by-step setup guides for every major AI coding and chat assistant.
Connect RewardSmart to Claude Desktop for conversational card recommendations using the native MCP config file.
Setup guideAdd RewardSmart as a custom connector in ChatGPT settings via the HTTP/SSE endpoint — no local install needed.
Setup guideAdd RewardSmart to Cursor's MCP config to get card advice while you code, right in the AI chat panel.
Setup guideUse RewardSmart with GitHub Copilot Chat in VS Code by adding an MCP config to your workspace or user settings.
Setup guideConnect RewardSmart to Windsurf's Cascade AI using the Codeium MCP config file for intelligent card guidance.
Setup guideIntegrate RewardSmart with Zed's AI assistant via the context_servers key in your Zed settings.json.
Setup guideAdd RewardSmart to Cline's MCP server list directly from the VS Code extension sidebar — no config file editing.
Setup guideWhat You Can Ask
Once connected, ask your AI assistant anything about your credit card rewards.