A secure Model Context Protocol server that enables AI agents to issue virtual cards, manage payment intents, and track expenses autonomously.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install proxy-pay-mcp
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install proxy-pay-mcp using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Proxy Pay MCP?
Proxy Pay MCP is a specialized integration designed to give AI agents a financial identity. By acting as a bridge between LLMs and payment infrastructure, it allows agents to perform complex financial tasks such as checking balances, simulating policy compliance, and creating payment intents. This skill is a core component for developers building autonomous procurement or expense management systems using Openclaw Skills.
The integration ensures security by separating agent capabilities from administrative functions. While agents can request funds and retrieve sensitive card details for authorized purchases, critical actions like funding accounts or final approval of large transactions remain restricted to human users, ensuring a robust human-in-the-loop governance model.
Proxy Pay MCP Use Cases
- Automating corporate procurement where agents buy software licenses or cloud credits within policy.
- Enabling autonomous travel agents to book flights and hotels with unique virtual cards.
- Streamlining expense management by having agents automatically attach receipts to transactions.
- Building interactive financial assistants in coding environments like Cursor or Claude via Openclaw Skills.
How Proxy Pay MCP Works
- The agent authenticates using either a permanent agent token for autonomous tasks or OAuth for interactive sessions.
- The system checks the KYC status and available balance to ensure the agent is authorized to spend.
- A policy simulation is performed to determine if the intended purchase requires human approval based on amount or merchant category.
- The agent creates a payment intent; if it exceeds auto-approval limits, a request_approval call is triggered for a human supervisor.
- Upon approval, the agent retrieves sensitive virtual card details linked specifically to that intent.
- After the purchase, the agent monitors the transaction list and attaches necessary receipts or evidence to close the loop.
Proxy Pay MCP Setup
To integrate this into your environment, add the server configuration to your MCP settings. For autonomous Openclaw Skills workflows, use the agent token method:
{
"mcpServers": {
"proxy": {
"type": "http",
"url": "https://mcp.useproxy.ai/api/mcp",
"headers": {
"Authorization": "Bearer $PROXY_AGENT_TOKEN"
}
}
}
}
For interactive use in IDEs, follow the OAuth login flow provided by the Proxy documentation.
Proxy Pay MCP Data Schema & Taxonomy
The skill organizes financial data through a structured hierarchy of intents and transaction metadata:
| Object | Description |
|---|---|
| Intent | The primary unit of a transaction; cards are locked until an intent is active. |
| Currency | All amounts are handled in cents (integer format) to avoid floating-point errors. |
| Policy | JSON-based rules defining autoApproveBelow limits and required approval flags. |
| Sensitive Data | PAN and CVV codes retrieved via intent-specific calls; never logged in plain text. |
| Evidence | Metadata and receipts linked to specific intent IDs for auditing. |
Proxy Pay MCP Advanced Features
- Policy Simulation: Allows agents to pre-check if a transaction will be blocked before attempting it.
- Human-in-the-Loop Safeguards: Hard-coded restrictions prevent agents from approving their own spending or closing accounts.
- Merchant MCC Analysis: Advisory tools to explain Merchant Category Codes and suggest allowlist entries.
- Dynamic Card Locking: Virtual cards remain inactive between authorized intents to minimize fraud risk within Openclaw Skills.
SKILL.md
Loading
Related Openclaw Skills
Proxy MCP Integration
A Model Context Protocol server that empowers AI agents to manage payments, issue virtual cards, and track financial transactions autonomously.
Proxy Balance Tracker
Monitor Proxy account balances and manage pending transaction intents in real-time.
AI Agent Card Payments
A secure framework for AI agents to autonomously manage virtual card payments through intent-based policy enforcement.
Agent Card Provisioning
Provision secure, policy-controlled virtual payment cards for AI agents to handle autonomous transactions.
Proxy Pay
A fintech automation skill for Openclaw Skills that enables instant creation of payment intents and provisioning of secure virtual cards via simple commands.
Proxy Pay MCP
A comprehensive Model Context Protocol skill that enables AI agents to autonomously request virtual payment cards and manage transactions within secure spending limits.
METADATA
Featured*
