# Agentpay Sentinel MCP

> Agentpay Sentinel MCP is a freemium MCP server for finance and payments. Pre-flight security checks for AI agent payments, returning PASS or BLOCKED per. It is available on CLI and API.

- **Kind:** MCP server
- **Pricing:** Freemium
- **Starting price:** $19 a month
- **Platforms:** CLI, API
- **Repository:** https://github.com/rumblingb/agentpay-sentinel-mcp
- **MCP Registry name:** io.github.Rumblingb/agentpay-sentinel-mcp
- **Categories:** [finance and payments](https://ai-powered.com/categories/finance-and-payments)
- **Listing updated:** 2026-11-18

## About

The README describes this server as a watchdog that an agent calls before it sends a payment. Each request runs through nine checks covering token hash integrity, budget caps, merchant allowlists, blocked categories, expiry, call rate, approved amount drift, nonce replay and revoked tokens. The verdict is advisory: the server does not sit on the network path, so the calling agent has to halt when it receives BLOCKED.

Beyond the audit tool, the server offers token revocation, nonce clearing for refunds, verification of audit hash chains, and a threat model tool that simulates named attack vectors such as token forgery, replay and merchant spoofing. Audit state is stored under a local home directory folder.

The project suits developers building agents that spend money on a user's behalf and want a logged, checkable decision before each purchase. The README says it is MIT licensed, with a free tier of 50 audits per month and a Pro plan at $19 per month for unlimited audits and threat model runs.

## Features

- **Nine-point payment audit:** Runs budget, allowlist, category, expiry, rate, amount, replay and revocation checks on each payment request.
- **Token integrity hashing:** Compares a SHA-256 hash against the expected merchant and amount binding to flag forged tokens.
- **Budget cap enforcement:** Rejects spends that would exceed the cap and reports the remaining budget.
- **Tamper-evident audit hashes:** Returns a SHA-256 digest of check results and can verify that a sequence of audit hashes forms a chain.
- **Token revocation:** Permanently blocks future audits for a revoked token ID.
- **Attack vector simulation:** Describes severity, defence and detection for named attack vectors.

## Good for

- Halt agent purchases that would exceed a set budget cap
- Confirm a payment amount still matches what a human approved
- Block replayed or revoked payment tokens before they execute
- Keep a verifiable record of pre-payment decisions for review

## Tools

- `sentinel_audit_transaction`: Runs all 9 security checks and returns PASS or BLOCKED with a SHA-256 audit hash.
- `sentinel_revoke_token`: Permanently revokes a token so future audits with its ID are blocked.
- `sentinel_clear_nonce`: Removes a nonce from the replay store for legitimate refunds or retries.
- `sentinel_verify_chain`: Verifies that a sequence of audit hashes forms an unbroken chain.
- `sentinel_threat_model`: Simulates a named attack vector and returns severity, defence and detection.

## FAQ

### Does the server block payments on its own?

No. The README says it returns an advisory verdict and does not intercept network traffic, so the agent must call it first and halt on BLOCKED.

### Which MCP clients does the README show setup for?

It gives configuration examples for Claude Desktop and Cursor, both pointing to a local server.py file.

### What does the free plan include?

The README lists 50 audits per month on the free plan. Unlimited audits and threat model simulations are part of the $19 per month Pro plan.

### Is Agentpay Sentinel MCP free?

Agentpay Sentinel MCP has a free plan, with paid plans for more usage or features. Paid plans start at $19 a month.

### What are the best alternatives to Agentpay Sentinel MCP?

The closest alternatives to Agentpay Sentinel MCP are SafePay Check, Agent Wallet MCP, Pink Agentic AI Payments, PayStage and Angel Agent Guard: pre-purchase check + prompt-injection scan.

## Where this comes from

- Description: written from the project's README and its MCP Registry entry.

## Alternatives

- [SafePay Check](https://ai-powered.com/mcp/safepay-check): SafePay Check is a freemium MCP server for finance and payments.
- [Agent Wallet MCP](https://ai-powered.com/mcp/agent-wallet-mcp): Agent Wallet MCP is a paid MCP server for finance and payments.
- [Pink Agentic AI Payments](https://ai-powered.com/mcp/pink-agentic-ai-payments): Pink Agentic AI Payments is a freemium MCP server for finance and payments.
- [PayStage](https://ai-powered.com/mcp/paystage-2): PayStage is an open-source MCP server for security.
- [Angel Agent Guard: pre-purchase check + prompt-injection scan](https://ai-powered.com/mcp/angel-agent-guard-pre-purchase-check-prompt-injection-scan): Angel Agent Guard: pre-purchase check + prompt-injection scan is a freemium MCP server for ai models and llm tooling.
- [BotsRated (beta)](https://ai-powered.com/mcp/botsrated-beta): BotsRated (beta) is a free MCP server for security.
- [PayStage](https://ai-powered.com/mcp/paystage): PayStage is an open-source MCP server for finance and payments.
- [XPayr Non-Custodial Crypto Gateway & AI Agent Commerce MCP](https://ai-powered.com/mcp/xpayr-non-custodial-crypto-gateway-and-ai-agent-commerce-mcp): XPayr Non-Custodial Crypto Gateway & AI Agent Commerce MCP is a paid MCP server for finance and payments.

---
Source: https://ai-powered.com/mcp/agentpay-sentinel-mcp
