gc-katana-mcp/build/index.js

Model Context Protocol (MCP) server for Katana - Fast web crawler with JavaScript parsing capabilities

#!/usr/bin/env node
import { McpServer } from "@modelcontextprotocol/sdk/server/mcp.js";
import { StdioServerTransport } from "@modelcontextprotocol/sdk/server/stdio.js";
import { z } from "zod";
import { spawn } from 'node:child_process';
// Get katana path from environment variable
const katanaPath = process.env.KATANA_PATH;
if (!katanaPath) {
    console.error("KATANA_PATH environment variable not set");
    process.exit(1);
}
// Create server instance
const server = new McpServer({
    name: "katana",
    version: "1.0.0",
});
function removeAnsiCodes(str) {
    // eslint-disable-next-line no-control-regex
    return str.replace(/\x1b\[[0-9;]*m/g, '');
}
server.tool("do-katana", "Performs fast and configurable web crawling on the given target URLs, identifying endpoints, parameters, and JS-based links.", {
    target: z.array(z.string()).describe("List of target URLs (e.g., https://example.com) to scan for endpoints and JavaScript-based links."),
    exclude: z.array(z.string()).optional().describe("List of URLs or regex patterns to exclude from crawling."),
    depth: z.number().optional().describe("Maximum crawl depth (e.g., 3 for three levels deep)."),
    js_crawl: z.boolean().optional().describe("Enable crawling and endpoint extraction from JavaScript files."),
    jsluice: z.boolean().optional().describe("Enable JSluice parsing for deeper JavaScript-based link analysis (memory intensive)."),
    headers: z.array(z.string()).optional().describe("List of custom headers or cookies to include in requests (format: Header:Value)."),
    strategy: z.enum(["depth-first", "breadth-first"]).optional().describe("Crawling strategy to use: 'depth-first' or 'breadth-first' (default is depth-first)."),
    headless: z.boolean().optional().describe("Enable headless browser-based hybrid crawling (experimental)."),
    system_chrome: z.boolean().optional().describe("Use the locally installed Chrome browser instead of the built-in one."),
    show_brwoser: z.boolean().optional().describe("Show the browser window even in headless mode (for debugging/visual inspection)."),
}, async ({ target, exclude, depth, js_crawl, jsluice, headers, strategy, headless, system_chrome, show_brwoser }) => {
    const katanaArgs = ["-u", target.join(","), "-silent"];
    if (exclude && exclude.length > 0) {
        katanaArgs.push("-exclude", exclude.join(","));
    }
    if (depth !== undefined) {
        katanaArgs.push("-d", depth.toString());
    }
    if (js_crawl) {
        katanaArgs.push("-jc");
    }
    if (jsluice) {
        katanaArgs.push("-jsl");
    }
    if (headers && headers.length > 0) {
        for (const header of headers) {
            katanaArgs.push("-H", header);
        }
    }
    if (strategy) {
        katanaArgs.push("-strategy", strategy);
    }
    if (headless) {
        katanaArgs.push("-headless");
    }
    if (system_chrome) {
        katanaArgs.push("-system-chrome");
    }
    if (show_brwoser) {
        katanaArgs.push("-show-browser");
    }
    let output = "";
    const katana = spawn(katanaPath, katanaArgs);
    // Handle stdout
    katana.stdout.on('data', (data) => {
        output += data.toString();
    });
    // Handle stderr
    katana.stderr.on('data', (data) => {
        output += data.toString();
    });
    // Handle process completion
    return new Promise((resolve, reject) => {
        katana.on('close', (code) => {
            if (code === 0 || typeof code === "undefined") {
                output = removeAnsiCodes(output);
                resolve({
                    content: [{
                            type: "text",
                            text: output
                        }]
                });
            }
            else {
                reject(new Error(`katana exited with code ${code}`));
            }
        });
        katana.on('error', (error) => {
            reject(new Error(`Error to start katana: ${error.message}`));
        });
    });
});
// Start the server
async function main() {
    const transport = new StdioServerTransport();
    await server.connect(transport);
    console.error("Katana MCP Server running on stdio");
}
main().catch((error) => {
    console.error("Fatal error in main():", error);
    process.exit(1);
});