Pane – Financial Data MCP Server for AI Agents

TL;DR

TL;DR: Pane is a hosted MCP server that connects your financial accounts (banks, cards, investments) to AI agents like Claude, ChatGPT, and Cursor — letting you ask natural questions about your money.

Source and Accuracy Notes

⚠️ This section is MANDATORY. All links must be verified from actual source, not guessed.

• Project page: pane.money
• HN launch thread: news.ycombinator.com/item?id=47240363
• License: Proprietary (SaaS, free tier available)
• Source last checked: 2026-08-12

What Is Pane?

Pane is a web service that acts as an MCP (Model Context Protocol) server for your financial data. Instead of giving an AI agent raw bank login credentials, you connect your accounts through Pane and get an MCP endpoint that any MCP-compatible client can query.

The product sits between your financial accounts and your AI tools. Supported account types include checking/savings, credit cards, and investment portfolios. Once linked, you can ask questions like:

• “What did I spend on food this month?”
• “What is my net worth right now?”
• “Show me my recurring subscriptions”
• “How much do I owe across all credit cards?”
• “What are my investment holdings?”

The service starts at $5/month and works with Claude (via the MCP protocol), ChatGPT (via OpenAI’s tools interface), Cursor, and any other MCP-compatible client.

Setup Workflow

Step 1: Create an Account

Sign up at pane.money. The free tier is available but limited; the $5/month Pro plan unlocks full account linking.

Step 2: Connect Your Financial Accounts

After logging in, navigate to Accounts and link your bank, card, and investment accounts. Pane uses read-only access — it cannot move money or initiate transactions.

Step 3: Configure Your MCP Client

Once your accounts are linked, Pane provides an MCP server endpoint. Configure your AI client to use it:

{
  "mcpServers": {
    "pane": {
      "command": "npx",
      "args": ["-y", "@pane-money/mcp-client"],
      "env": {
        "PANE_API_KEY": "your-api-key-here"
      }
    }
  }
}

Step 4: Start Chatting

With the MCP server configured, your AI agent can now reason about your financial context:

User: How much did I spend on dining out last month?
Agent (via Pane MCP): You spent $347.82 on dining out in July,
across 23 transactions at restaurants including $142 at full-service
establishments and $205 at fast casual spots.

How Pane Works Under the Hood

Pane operates as a hosted MCP server rather than a local binary. When you query your financial data:

1. Your AI client sends a request through the MCP protocol
2. Pane authenticates via your API key and checks account permissions
3. Pane queries aggregated data from your linked financial accounts
4. Structured financial context is returned to the AI agent

This differs from giving an AI your bank login directly. Pane normalizes data across account types and provides a clean MCP interface instead of screen-scraping a banking website.

Practical Evaluation Checklist

• Read-only access across all linked accounts
• MCP protocol compatible with Claude, ChatGPT, Cursor, and any MCP client
• Normalizes data across bank, card, and investment account types
• API key authentication (no bank credentials stored by Pane)
• Starting at $5/month for full access

Security Notes

Pane connects to financial institutions using read-only aggregator services. The MCP server stores only an API key — not bank login credentials. All data is transmitted over HTTPS through the MCP protocol.

As with any tool that gives an AI agent access to personal data, consider the threat model: a compromised AI agent with a Pane API key could read your financial data but cannot initiate transactions. Rotate API keys through the Pane dashboard if you suspect compromise.

FAQ

Q: Can Pane move money or make transactions? A: No. Pane provides read-only access to your financial data. It cannot initiate transfers, payments, or any transaction.

Q: Which financial institutions are supported? A: Pane supports major US banks, credit card issuers, and brokerages. The product page lists specific partners. International support depends on the aggregator coverage.

Q: Does Pane work with local AI models? A: Any AI client that implements the MCP protocol can connect to Pane, including locally running models with MCP tool-calling support.

Q: What happens to my data if I cancel? A: Your linked accounts are disconnected and API keys are revoked. Pane’s data retention policy is available in their privacy policy.

Conclusion

Pane fills a specific niche: connecting AI agents to structured financial context without handing over bank credentials. At $5/month, it is cost-competitive with standalone personal finance trackers, while adding MCP protocol access that fits directly into an AI-first workflow. For users who want AI agents that understand their complete financial picture — not just a document or database — Pane provides a clean integration point.