Codesafer

Overview

Codesafer is a Model Context Protocol server that scans AI-generated code for invisible Unicode characters, Trojan Source attacks, and supply-chain threats. Any MCP-compatible AI assistant can invoke Codesafer's analysis tools directly during conversations or agent workflows, letting the model decide when to apply security checks to generated code.

The server runs locally via the npm package cleaner-code, which communicates over stdio without requiring any credentials or external authentication. Version v1.0.1 is available under the MIT license, with full source code at goldmembrane/cleaner-code on GitHub.

Codesafer belongs in the Development category on mcp.site and integrates with MCP clients including Claude Desktop, Claude Code, Cursor, VS Code, Windsurf, and Zed. If you maintain this project, claim the listing to verify ownership, receive the Verified badge, and ensure information stays accurate.

Installation

// claude_desktop_config.json
{
  "mcpServers": {
    "codesafer": {
      "command": "npx",
      "args": [
        "-y",
        "cleaner-code"
      ]
    }
  }
}