Skip to main content
Glama
zbkm

mmnt-mcp-server

by zbkm

mmnt_cache

Extract cached web pages from the Mamont search engine by providing a unique cache ID, with options for text-only or HTML output.

Instructions

Extract page from Mamont cache

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
idYesunique cache id
onlyTextYesShould the result be text only (no html)

Implementation Reference

  • The core handler function that fetches the cached HTML page from mmnt.ru, decodes from windows-1251 to UTF-8, and extracts plain text from the <pre> element if onlyText is true.
    export async function extractCache(id: string, onlyText: boolean = true): Promise<string> {
        const response = await fetch(`https://www.mmnt.ru/cache2/${id}.html`);
        const buffer = await response.arrayBuffer();
        const html = iconv.decode(Buffer.from(buffer), "windows-1251"); // convert win1251 -> utf8
    
        if (!onlyText) {
            return html;
        }
    
        const $ = cheerio.load(html);
        return $("pre").first().text();
    }
  • src/index.ts:27-37 (registration)
    Registers the 'mmnt_cache' tool with the MCP server, using CacheParams schema and wrapping the extractCache handler.
    server.tool(
        "mmnt_cache",
        "Extract page from Mamont cache",
        CacheParams,
        async ({id, onlyText}) => ({
            content: [{
                type: "text",
                text: JSON.stringify(await extractCache(id, onlyText))
            }]
        })
    );
  • Zod schema defining input parameters for the mmnt_cache tool: id (string) and onlyText (boolean).
    export const CacheParams = {
        id: z.string({description: "unique cache id"}),
        onlyText: z.boolean({description: "Should the result be text only (no html)"})
    };
Install Server

Other Tools

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/zbkm/mmnt-mcp-server'

If you have feedback or need assistance with the MCP directory API, please join our Discord server