AgentCard

Created By
Karen Serfaty3 months ago
Virtual debit cards for AI agents. Create, spend, and manage budget-locked Mastercard debit cards through MCP. Each card is single-use, encrypted, and capped at its funded amount — no overdraft possible.
Overview

AgentCard MCP Server

Give your AI agent a debit card. AgentCard lets AI agents create and spend virtual Mastercard debit cards — each with a fixed budget, real card credentials, and full MCP integration.

Features

  • MCP-native: One config block connects to Claude Code, Claude Desktop, Cursor, or any MCP client
  • Human-in-the-loop: Card funding requires approval via Stripe Checkout
  • Single-use by design: Cards are disposable with fixed limits, no overdraft possible
  • AES-256-GCM encryption: Card credentials encrypted at rest per-card
  • Instant issuance: Cards ready in seconds after funding

Quick Setup

npx agent-cards signup       # Create account (one-time)
npx agent-cards setup-mcp    # Auto-configures Claude Code

Restart Claude Code and your agent can create and spend virtual cards.

Manual Setup (HTTP)

{
  "mcpServers": {
    "agent-cards": {
      "url": "https://mcp.agentcard.sh/mcp",
      "headers": {
        "Authorization": "Bearer <your-jwt>"
      }
    }
  }
}

Get your JWT with: npx agent-cards login

Available Tools

Card Management

  • list_cards — List all virtual cards with IDs, last 4 digits, expiry, balance, and status
  • create_card — Create a new virtual debit card with specified USD amount ($1-$50)
  • get_card_details — Get decrypted PAN, CVV, expiry date, and current balance
  • check_balance — Check live balance without exposing credentials
  • close_card — Permanently close a virtual card

Checkout Integration

  • pay_checkout — Auto-fill a checkout page with an AgentCard via Chrome extension
  • detect_checkout — Detect if current browser tab is a checkout page
  • fill_card — Fill an existing card into the current checkout form

Payment Setup

  • setup_payment_method — Set up a saved payment method via Stripe Checkout
  • submit_user_info — Submit identity information required before first card creation

Transaction Tracking

  • list_transactions — List transactions for a card with merchant, amount, status, timestamps

Approval Flow

  • approve_request — Resolve pending approval requests for card creation

Support

  • start_support_chat — Start a new support conversation
  • send_support_message — Send a message in an existing support conversation
  • read_support_chat — Read message history of a support conversation

How It Works

  1. Developer installs: npx agent-cards signup
  2. Sets up MCP: npx agent-cards setup-mcp
  3. Agent calls create_card with USD amount
  4. Human approves funding via Stripe Checkout
  5. Agent gets card details with get_card_details
  6. Agent uses card to pay for services, APIs, or purchases
  7. Card is auto-limited to funded amount — no overdraft possible

Server Config

{
  "mcpServers": {
    "agent-cards": {
      "url": "https://mcp.agentcard.sh/mcp",
      "headers": {
        "Authorization": "Bearer <your-jwt>"
      }
    }
  }
}
Project Info
Created At
3 months ago
Updated At
3 months ago
Author Name
Karen Serfaty
Star
-
Language
-
License
-
Category

Recommend Servers

View All
Bring your real authenticated browser session to AI coding agents. Local-first MCP server + Chrome MV3 extension. No cloud. No telemetry.
@Cubenest

peek records the user's actual logged-in browser (DOM via rrweb, console events, network metadata, optional response bodies via opt-in Deep capture) through a Chrome MV3 extension. The extension ships events through a native-messaging stdio bridge to a local MCP server (peek-mcp), which persists them to a SQLite database at ~/.peek/sessions.db. AI coding agents (Claude Code, Cursor, Cline, Windsurf) read sessions from the database via 10 MCP tools: Tool What it does list_recent_sessions List recently recorded sessions (id, origin, ts, event count). get_session_summary LLM-readable narrative summary of a session. get_session_console_errors Console errors recorded in a session. get_session_network_errors Failed/notable network requests in a session. get_user_action_before_error Last N user actions before a console error. generate_playwright_repro Generate a runnable Playwright test from a session. get_dom_snapshot Reconstruct the DOM at a given timestamp. query_dom_history Timeline of attribute/text changes for a selector. request_authorization Side-panel consent for write actions (Level 3). execute_action Dispatch a UI action (gated by permission level + destructive blocklist). Why local-first matters Every other "browser session for AI" tool ships to a vendor cloud. peek's SQLite + extension live on the user's machine — no remote endpoints, no telemetry. The privacy policy (docs/peek/PRIVACY_POLICY.md) is the source of truth. Install # 1. Add the MCP server to Claude Code claude mcp add peek -- npx -y @peekdev/mcp # 2. Install the Chrome extension from the Chrome Web Store # (link added once the CWS listing is approved)

a day ago