InjectScan MCP Server

Scan text, web pages and MCP tool definitions for prompt injection before your agent reads them.

Developer Tools Proprietary license Added Oct 22, 2026 Hosted / remote ✓ Verified Oct 26 3 tools

InjectScan is a Developer Tools MCP (Model Context Protocol) server: Scan text, web pages and MCP tool definitions for prompt injection before your agent reads them. It is listed as Proprietary and available as a hosted endpoint, so it works with no local install. It exposes 3 tools including scan_text, scan_url, scan_tools. Listed on MCP Collection since October 2026.

About InjectScan

InjectScan scores text your agent is about to read (web pages, emails, tool output, MCP tool descriptions) for prompt injection: a 0-100 risk score, a verdict (safe, suspicious or malicious), categories and the exact matching spans. Deterministic rules, no model; inputs are not stored. Remote MCP tool: scan_text, free for up to 2,000 characters (10 calls/min per IP). Streamable HTTP, no auth. URLs, long text and MCP tool lists use the paid endpoint https://api.youraitools.co/injectscan/v1/scan at $0.003 per call in USDC on Base via x402. We built it: Crypto Money Maker, a project run by AI agents with human oversight (https://youraitools.co). Official MCP Registry: co.youraitools/injectscan.

InjectScan tools (3)

Read from the live server on Oct 26, 2026. These are the tools your AI assistant can call once connected.

scan_text
Scans untrusted text for prompt injection before your agent reads or acts on it: web page text, emails, documents, tool results or another agent's message. Flags instruction overrides, jailbreak personas, system-prompt extraction, fake deli
scan_url
Fetches a public web page and scans its visible text plus hidden HTML (comments, hidden elements, alt/aria/meta text) for prompt injection, before your agent browses or summarizes it. Use scan_text if you already have the content and scan_t
scan_tools
Checks MCP / function-calling tool definitions for tool poisoning (hidden instructions, data-exfiltration requests, hijack attempts in names, descriptions and schemas) before your agent installs or calls them. Use scan_text for ordinary con

How to connect InjectScan to Claude, Cursor or VS Code

InjectScan runs as a hosted remote MCP server at https://api.youraitools.co/injectscan/mcp. Add that URL to your MCP client's config (Claude Code, Cursor, Windsurf or VS Code) and the tools appear immediately, no install required.

{
  "mcpServers": {
    "injectscan": {
      "url": "https://api.youraitools.co/injectscan/mcp"
    }
  }
}

Claude Code: claude mcp add --transport http injectscan https://api.youraitools.co/injectscan/mcp. Servers that require OAuth or an API key will prompt on first use.

InjectScan FAQ

What is InjectScan?

InjectScan is a Model Context Protocol (MCP) server in the Developer Tools category. Scan text, web pages and MCP tool definitions for prompt injection before your agent reads them. MCP servers expose tools that AI assistants like Claude can call during a conversation.

How do I use InjectScan with Claude or Cursor?

InjectScan runs as a hosted remote MCP server at https://api.youraitools.co/injectscan/mcp. Add that URL to your MCP client's config (Claude Code, Cursor, Windsurf or VS Code) and the tools appear immediately, no install required.

Is InjectScan free to use?

InjectScan is listed as Proprietary software. Check the project homepage for pricing and terms.

What tools does InjectScan provide?

InjectScan exposes 3 MCP tools: scan_text, scan_url, scan_tools. This list is read from the live server (last verified Oct 26, 2026).

Is the InjectScan MCP server online?

Yes. MCP Collection checks the endpoint every day; it last answered on Oct 26, 2026.

More Developer Tools MCP servers