PromptGuard Scan MCP Server

Remote MCP for prompt injection scans, jailbreak checks, leakage review, and LLM release gates.

GitHub

Documentation

PromptGuard Scan MCP

PromptGuard Scan MCP is a paid remote Model Context Protocol server for LLM security testing. It helps teams inspect AI apps for prompt injection, jailbreaks, hidden instruction leakage, unsafe tool-use signals, and release-gate risk.

This repository is the public MCP integration and directory listing package for the hosted service at promptguardscan.space. The production MCP endpoint is hosted remotely; the product source code is not published here.

MCP endpoint

Tokens are issued after checkout and MCP token claim. Do not paste access tokens into public issues, prompts, screenshots, logs, or pull requests.

Connect

Use your issued Bearer token in an MCP client that supports remote Streamable HTTP servers.

{
  "mcpServers": {
    "promptguardscan": {
      "type": "http",
      "url": "https://promptguardscan.space/mcp",
      "headers": {
        "Authorization": "Bearer <token>"
      }
    }
  }
}

What it is for

  • Prompt injection scanner workflows for AI apps, chatbots, agents, RAG systems, and custom GPT-style tools.
  • Jailbreak detection and prompt security regression testing before release.
  • LLM vulnerability scanner evidence for engineering teams and buyer security reviews.
  • CI/CD release-gate context for high-risk prompt, retrieval, and tool-use changes.

Public directory records

  • Official MCP Registry name: space.promptguardscan/promptguardscan-mcp
  • Smithery qualified name: clauxel/prompt-injection-scanner-mcp
  • Primary website: https://promptguardscan.space/

Related security guides

Support

For account, checkout, token, or security-report questions, contact [email protected].

Related Workflow

  • OpenHuman Online is useful when this MCP rollout also needs human-readable source memory, meeting notes, or approval context outside the server.