co.promptguard/security

Official

PromptGuard

Scan prompts for injection attacks, redact PII, and audit LLM SDK usage from any MCP client

Detect prompt injection attacks, redact PII, and audit LLM SDK usage.

Install

Configuration

Environment variables

PROMPTGUARD_API_KEYsecret

PromptGuard API key (starts with pg_sk_test_ or pg_sk_prod_)

PROMPTGUARD_API_URL

API base URL for self-hosted deployments

claude_desktop_config.json

{
  "mcpServers": {
    "security": {
      "command": "uvx",
      "args": [
        "promptguard-mcp-server"
      ],
      "env": {
        "PROMPTGUARD_API_KEY": "<YOUR_PROMPTGUARD_API_KEY>",
        "PROMPTGUARD_API_URL": "<YOUR_PROMPTGUARD_API_URL>"
      }
    }
  }
}