{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": "# Quickstart: Toolroll: Official MCP Registry Snapshot (every MCP server, weekly)\n\nA snapshot of the official Model Context Protocol (MCP) Registry (registry.modelcontextprotocol.io) as a single Parquet table: 35,403 MCP servers (35,028 active, 375 deprecated) from 20,000+ publishers, snapshot 2026-09-24 07:03 UTC.\n\nDataset: [huggingface.co/datasets/CyberMax-tools/toolroll-mcp-registry-snapshot](https://huggingface.co/datasets/CyberMax-tools/toolroll-mcp-registry-snapshot) · file: `data/servers.parquet` · by CyberMax.\n\nRuns anywhere with pandas (Colab, Kaggle, Jupyter, VS Code). The outputs below are from a real run of this notebook."
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": "## Load the data"
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": "35,403 rows x 21 columns\n"
    },
    {
     "output_type": "execute_result",
     "execution_count": 1,
     "metadata": {},
     "data": {
      "text/plain": "                                     name  ... needs_secret\n0                     ac.inference.sh/mcp  ...        False\n1                            ac.snag/snag  ...        False\n2                      ac.tandem/docs-mcp  ...        False\n3                    ad.inside/inside-ads  ...        False\n4                      ae.propick/propick  ...         True\n5                    ag.hood/name-service  ...        False\n6                        agency.goji/goji  ...        False\n7                    agency.kesey/pretrip  ...        False\n8                     agency.lona/trading  ...        False\n9  agency.ottobot/business-contact-finder  ...        False\n\n[10 rows x 21 columns]",
      "text/html": "<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>name</th>\n      <th>namespace</th>\n      <th>title</th>\n      <th>description</th>\n      <th>version</th>\n      <th>status</th>\n      <th>...</th>\n      <th>remote_types</th>\n      <th>remote_urls</th>\n      <th>has_remote</th>\n      <th>has_package</th>\n      <th>env_var_count</th>\n      <th>needs_secret</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>ac.inference.sh/mcp</td>\n      <td>ac.inference.sh</td>\n      <td>inference.sh</td>\n      <td>run any ai model. compose agents, stack knowledge, connect tools. one api, pay per run.</td>\n      <td>2.0.1</td>\n      <td>active</td>\n      <td>...</td>\n      <td>[streamable-http]</td>\n      <td>[https://api.inference.sh/mcp]</td>\n      <td>True</td>\n      <td>False</td>\n      <td>0</td>\n      <td>False</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>ac.snag/snag</td>\n      <td>ac.snag</td>\n      <td></td>\n      <td>Gives your coding agent the captured console, network, replay and screenshot evidence for a bug.</td>\n      <td>1.0.0</td>\n      <td>active</td>\n      <td>...</td>\n      <td>[streamable-http]</td>\n      <td>[https://mcp.snag.ac/mcp]</td>\n      <td>True</td>\n      <td>False</td>\n      <td>0</td>\n      <td>False</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>ac.tandem/docs-mcp</td>\n      <td>ac.tandem</td>\n      <td></td>\n      <td>Remote MCP server for Tandem docs, install guides, SDKs, workflows, and agent setup help.</td>\n      <td>0.3.2</td>\n      <td>active</td>\n      <td>...</td>\n      <td>[streamable-http]</td>\n      <td>[https://tandem.ac/mcp]</td>\n      <td>True</td>\n      <td>False</td>\n      <td>0</td>\n      <td>False</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>ad.inside/inside-ads</td>\n      <td>ad.inside</td>\n      <td>Inside Ads</td>\n      <td>Telegram ad exchange: estimate reach and cost with no account, then create and run campaigns.</td>\n      <td>1.0.0</td>\n      <td>active</td>\n      <td>...</td>\n      <td>[streamable-http]</td>\n      <td>[https://app.inside.ad/api/mcp]</td>\n      <td>True</td>\n      <td>False</td>\n      <td>0</td>\n      <td>False</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>ae.propick/propick</td>\n      <td>ae.propick</td>\n      <td>Propick Integration MCP</td>\n      <td>Manage your real-estate stock on Propick (Dubai): bulk listing sync, lookups and run reports.</td>\n      <td>1.0.0</td>\n      <td>active</td>\n      <td>...</td>\n      <td>[streamable-http]</td>\n      <td>[https://propick.ae/mcp]</td>\n      <td>True</td>\n      <td>False</td>\n      <td>0</td>\n      <td>True</td>\n    </tr>\n    <tr>\n      <th>5</th>\n      <td>ag.hood/name-service</td>\n      <td>ag.hood</td>\n      <td>hood. — .hood name service</td>\n      <td>Resolve .hood names on Robinhood Chain — forward/reverse, text records, availability &amp; pricing.</td>\n      <td>0.1.0</td>\n      <td>active</td>\n      <td>...</td>\n      <td>[streamable-http]</td>\n      <td>[https://www.hood.ag/api/mcp]</td>\n      <td>True</td>\n      <td>False</td>\n      <td>0</td>\n      <td>False</td>\n    </tr>\n    <tr>\n      <th>6</th>\n      <td>agency.goji/goji</td>\n      <td>agency.goji</td>\n      <td></td>\n      <td>Answers on AEO, SEO, web and brand from GOJI's published material. Melbourne, Australia.</td>\n      <td>1.0.1</td>\n      <td>active</td>\n      <td>...</td>\n      <td>[streamable-http]</td>\n      <td>[https://mcp.goji.agency/mcp]</td>\n      <td>True</td>\n      <td>False</td>\n      <td>0</td>\n      <td>False</td>\n    </tr>\n    <tr>\n      <th>7</th>\n      <td>agency.kesey/pretrip</td>\n      <td>agency.kesey</td>\n      <td>Pre-Trip compliance scanner</td>\n      <td>Screen regulated-health marketing copy against source-cited rulesets, all 50 states.</td>\n      <td>1.0.1</td>\n      <td>active</td>\n      <td>...</td>\n      <td>[]</td>\n      <td>[]</td>\n      <td>False</td>\n      <td>True</td>\n      <td>0</td>\n      <td>False</td>\n    </tr>\n    <tr>\n      <th>8</th>\n      <td>agency.lona/trading</td>\n      <td>agency.lona</td>\n      <td></td>\n      <td>AI-powered trading strategy development: backtesting, market data, and portfolio analysis</td>\n      <td>2.0.0</td>\n      <td>active</td>\n      <td>...</td>\n      <td>[streamable-http]</td>\n      <td>[https://mcp.lona.agency/mcp]</td>\n      <td>True</td>\n      <td>False</td>\n      <td>0</td>\n      <td>False</td>\n    </tr>\n    <tr>\n      <th>9</th>\n      <td>agency.ottobot/business-contact-finder</td>\n      <td>agency.ottobot</td>\n      <td>Business Contact Finder</td>\n      <td>Check how to contact a business website, and whether that contact path actually works.</td>\n      <td>0.2.0</td>\n      <td>active</td>\n      <td>...</td>\n      <td>[streamable-http]</td>\n      <td>[https://business-contact-finder-mcp.ottobot2025.workers.dev/mcp]</td>\n      <td>True</td>\n      <td>False</td>\n      <td>0</td>\n      <td>False</td>\n    </tr>\n  </tbody>\n</table>"
     }
    }
   ],
   "source": "import pandas as pd\n\nURL = \"https://huggingface.co/datasets/CyberMax-tools/toolroll-mcp-registry-snapshot/resolve/main/data/servers.parquet\"\ndf = pd.read_parquet(URL)\nprint(f\"{len(df):,} rows x {len(df.columns)} columns\")\ndf.head(10)"
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": "Rows per month by `published_at`:"
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 2,
     "metadata": {},
     "data": {
      "text/plain": "published_at\n2025-10      229\n2025-11      143\n2025-12      213\n2026-01      296\n2026-02      913\n2026-03     1573\n2026-04     1894\n2026-05     2440\n2026-06     3284\n2026-07     4537\n2026-08     7228\n2026-09    12305\nFreq: M, Name: count, dtype: int64"
     }
    }
   ],
   "source": "months = pd.to_datetime(df['published_at'], errors='coerce', utc=True).dt.tz_localize(None).dt.to_period('M')\nmonths.value_counts().sort_index().tail(12)"
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": "Summary statistics for the numeric columns:"
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 3,
     "metadata": {},
     "data": {
      "text/plain": "       env_var_count\ncount      35403.000\nmean           0.725\nstd            3.718\nmin            0.000\n25%            0.000\n50%            0.000\n75%            0.000\nmax          287.000",
      "text/html": "<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>env_var_count</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>count</th>\n      <td>35403.000</td>\n    </tr>\n    <tr>\n      <th>mean</th>\n      <td>0.725</td>\n    </tr>\n    <tr>\n      <th>std</th>\n      <td>3.718</td>\n    </tr>\n    <tr>\n      <th>min</th>\n      <td>0.000</td>\n    </tr>\n    <tr>\n      <th>25%</th>\n      <td>0.000</td>\n    </tr>\n    <tr>\n      <th>50%</th>\n      <td>0.000</td>\n    </tr>\n    <tr>\n      <th>75%</th>\n      <td>0.000</td>\n    </tr>\n    <tr>\n      <th>max</th>\n      <td>287.000</td>\n    </tr>\n  </tbody>\n</table>"
     }
    }
   ],
   "source": "df[['env_var_count']].describe().round(3)"
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": "## Go further\n- **All CyberMax tools, datasets and free apps:** [https://cybermax-tools-cybermax.static.hf.space/](https://cybermax-tools-cybermax.static.hf.space/)"
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "name": "python"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}