# Agent Guard Prompt Injection & PII Firewall (x402)

> Agent Guard Prompt Injection & PII Firewall (x402) is a paid MCP server for ai models and llm tooling. MCP server that screens prompts for injection and masks Brazilian PII via x402 USDC. It runs as a hosted remote server, so there is nothing to install locally.

- **Kind:** MCP server
- **Pricing:** Paid
- **Platforms:** API, CLI
- **Repository:** https://github.com/brunopompeo1/brazil-ai-fiscal-apis-x402
- **MCP Registry name:** io.github.brunopompeo1/agent-guard-x402
- **Categories:** [AI models and LLM tooling](https://ai-powered.com/categories/ai-and-llm), [security](https://ai-powered.com/categories/security)
- **Listing updated:** 2026-11-19

## About

The README describes this as one of seven pay-per-call APIs in a Brazilian tax and finance suite, here offered as a standalone guard service. Each call inspects text before it reaches a language model, combining the Llama Guard 3 classifier with heuristic checks and Base64 decoding to flag injection attempts.

Alongside injection screening, the service masks sensitive Brazilian identifiers such as CPF, CNPJ and key-like strings. The MCP endpoint is a streamable-HTTP remote, and the README notes that adding the URL alone does not set up automatic payment; the client must handle the x402 charge or send an X-API-Key header.

It suits developers building agents or automations that handle Brazilian customer or accounting documents and want a filter in front of the model. The README also says a free /sample route exists to inspect payloads before paying.

## Features

- **Prompt injection detection:** Flags injection attempts using Llama Guard 3 combined with heuristic rules and Base64 decoding.
- **Brazilian PII masking:** Masks CPF, CNPJ and key-like secrets before text is sent to a model.
- **Pay-per-call pricing:** Each guard call is priced at a small USDC amount on the Base network via x402.
- **Free sample endpoint:** A /sample route lets developers inspect the payload shape without paying.
- **Streamable HTTP MCP remote:** Exposes the service to MCP clients over a streamable-HTTP endpoint.

## Good for

- Screen user prompts for injection before forwarding them to an LLM agent
- Mask CPF and CNPJ numbers in text before logging or sending it to third-party models
- Add a pre-model safety filter to n8n or Make automations handling Brazilian documents

## Install

### Claude Code (Terminal)

```sh
claude mcp add --transport http agent-guard-x402 https://agent-guard-x402.architect-x402.workers.dev/mcp
```

### Claude Desktop (Settings → Connectors → Add custom connector)

```sh
https://agent-guard-x402.architect-x402.workers.dev/mcp
```

### Cursor (~/.cursor/mcp.json)

```json
{
  "mcpServers": {
    "agent-guard-x402": {
      "url": "https://agent-guard-x402.architect-x402.workers.dev/mcp"
    }
  }
}
```

### VS Code (.vscode/mcp.json)

```json
{
  "servers": {
    "agent-guard-x402": {
      "type": "http",
      "url": "https://agent-guard-x402.architect-x402.workers.dev/mcp"
    }
  }
}
```

### Windsurf (~/.codeium/windsurf/mcp_config.json)

```json
{
  "mcpServers": {
    "agent-guard-x402": {
      "serverUrl": "https://agent-guard-x402.architect-x402.workers.dev/mcp"
    }
  }
}
```

### Grok Bot (Grok Bot desktop app → Connect apps → Add a custom connector)

```sh
https://agent-guard-x402.architect-x402.workers.dev/mcp
```

Enter a Name (for example "Agent Guard Prompt Injection & PII Firewall (x402)"), choose Type: Remote HTTPS, paste this as the Server URL and choose Add plugin.

### Other clients (your client's MCP settings (most use this shape))

```json
{
  "mcpServers": {
    "agent-guard-x402": {
      "type": "http",
      "url": "https://agent-guard-x402.architect-x402.workers.dev/mcp"
    }
  }
}
```

## FAQ

### How much does a guard call cost?

The README lists $0.008 per call for the guard endpoint, paid in USDC on Base.

### Do I need a wallet to try it?

The README says discovery works without credentials, but real calls require either an X402_PRIVATE_KEY wallet with USDC on Base or a B2B API key.

### Does adding the MCP URL configure payment automatically?

No. The README states that adding only the remote URL does not set up automatic payment; the client must handle x402 or send X-API-Key.

### Is Agent Guard Prompt Injection & PII Firewall (x402) free?

No. Agent Guard Prompt Injection & PII Firewall (x402) is a paid product.

### How do I add Agent Guard Prompt Injection & PII Firewall (x402) to Claude Code?

Run this command in your terminal: claude mcp add --transport http agent-guard-x402 https://agent-guard-x402.architect-x402.workers.dev/mcp

### What are the best alternatives to Agent Guard Prompt Injection & PII Firewall (x402)?

The closest alternatives to Agent Guard Prompt Injection & PII Firewall (x402) are PromptGuard, APEX Agent Safety & Guardrails, CanaryGuard 402, Angel Agent Guard: pre-purchase check + prompt-injection scan and PentestLLM.

## Where this comes from

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

## Alternatives

- [PromptGuard](https://ai-powered.com/mcp/promptguard): PromptGuard is a freemium MCP server for security.
- [APEX Agent Safety & Guardrails](https://ai-powered.com/mcp/apex-agent-safety-and-guardrails): APEX Agent Safety & Guardrails is a paid MCP server for security.
- [CanaryGuard 402](https://ai-powered.com/mcp/canaryguard-402): CanaryGuard 402 is a paid 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.
- [PentestLLM](https://ai-powered.com/mcp/pentestllm): PentestLLM is a paid MCP server for security.
- [Guard Brasil](https://ai-powered.com/mcp/guard-brasil): Guard Brasil is an open-source MCP server for security.
- [Power Strip Security](https://ai-powered.com/mcp/power-strip-security): Power Strip Security is a paid MCP server for security.
- [Tinplate Edge AI Safety, Security & EU Compliance](https://ai-powered.com/mcp/tinplate-edge-ai-safety-security-and-eu-compliance): Tinplate Edge AI Safety, Security & EU Compliance is a freemium MCP server for security.

---
Source: https://ai-powered.com/mcp/agent-guard-prompt-injection-and-pii-firewall-x402
