mcpField
G
grouppulse

grouppulse

Local
github.com/CloudCorpRecordsCommunication0Updated 1w ago

About grouppulse

grouppulse is an MCP server for read-only statistics from public Telegram groups, intended for people exploring group activity or comparing communities. It provides group search, per-group activity and member summaries, and network-wide totals through the Limzo public API.

Key features

  1. 01Search the public group directory
  2. 02View activity and member rankings
  3. 03Inspect levels, awards, and moderation
  4. 04Compare network-wide group statistics

Use cases

  1. 01Find public groups by topic or language
  2. 02Review a group's activity over time
  3. 03Inspect member rankings and award history
  4. 04Check overall public Telegram group trends

MCP support

Exposes three read-only tools to MCP clients: search public groups, retrieve a group's statistics, and view network-wide totals.

Configuration

Local server config is the same for every client. Copy it into your client’s MCP config file.

mcp.json · Universal config
{
"mcpServers": {
"cloudcorprecords-grouppulse": {
"command": "npx",
"args": [
"-y",
"@reneturcios/grouppulse"
]
}
}
}
Run commandnpx -y @reneturcios/grouppulse

Local servers need a runtime on your machine (e.g. Node.js 18+). Config file locations differ by client; see your client’s docs.

Packages
npm@reneturcios/grouppulse0.1.0
ocighcr.io/cloudcorprecords/grouppulse:0.1.0

Tools

3

Saved tool list · updated 1w ago

ParameterTypeRequiredDescription
limitintegerNoMaximum number of results to return (optional).
querystringNoSearch query text.
languagestringNoLanguage code (optional).

Details

From the project README

grouppulse

Public Telegram group statistics, with no key to set up. grouppulse talks to the Limzo public API, which publishes read-only aggregate statistics for public Telegram groups: how active a group is, who its most active members are, how its level ladder looks, what awards it has handed out, and a moderation summary. Nothing here needs an account, and the upstream data never contains verbatim member messages — only aggregates and leaderboards — so what grouppulse passes through inherits that property.

Three questions cover most of what the data can answer: which public groups exist around a topic or a language, how one particular group is doing, and how the whole network looks in total.

Configuration

  • LIMZO_BASE_URL (optional) — API base URL. Defaults to https://limzo.com.

That is all. There is no API key: every endpoint is public and read-only.

Tools

  • list_groups — search or browse the public group directory. Each entry carries its slug, title, member count and language. All arguments are optional: query (a keyword matched against group titles, usernames, slugs and descriptions), language (an ISO 639-1 code such as fa or es), and limit (how many groups to return). Omit everything to list the top groups.
  • get_group_stats — full public stats for one group, by its slug (the last part of its limzo.com/s/<slug> page). The result covers activity, top members, levels, awards and a moderation summary, plus the group's mood and language mix when the upstream has enough signal. The optional range picks the stats window: 7d (the default), 30d or all; the longer windows need the group's paid plan upstream, and on a free group the answer silently falls back to 7d, which you can spot in the returned range block.
  • get_global_stats — network-wide totals across all public groups, plus the current top groups.

Behaviour worth knowing:

  • A 2xx answer comes back exactly as the API returned it: nothing is trimmed, renamed or summarised on the way through.
  • Any other reply fails with grouppulse error: HTTP <status> — for example a 404 for an unknown slug, a 403 when a group's public page is disabled, a 410 when its stats were deleted, or a 429 when you poll faster than the upstream allows (5 requests per second). An unreachable API fails with grouppulse error: unreachable. These error strings are this rebuild's own.
  • The upstream asks clients to respect its cache headers (responses are cached for 30 minutes); stats move slowly, so there is rarely a reason to poll.
Endpoint provenance

The vendor's published OpenAPI document pins the endpoints, the wire parameter names and the range values; the original MCP project's README pins the three tools. The MCP-layer argument names are labelled reconstructions mapped onto the pinned wire parameters:

  • GET <base>/api/public/groups with q, lang and limit when set (used by list_groups; the MCP arguments are query, language and limit).
  • GET <base>/api/public/{slug} with range when set (used by get_group_stats; the /s/{slug}.json page route carries the same payload, and the /api/public/ path is the API route).
  • GET <base>/api/public/global-stats (used by get_global_stats).

The slug path segment is URL-encoded.

CLI

Each tool is a subcommand taking one JSON argument string:

npx @reneturcios/grouppulse list_groups '{"language":"es","limit":5}'
npx @reneturcios/grouppulse get_group_stats '{"slug":"hipo"}'
npx @reneturcios/grouppulse get_group_stats '{"slug":"hipo","range":"30d"}'
npx @reneturcios/grouppulse get_global_stats '{}'
npx @reneturcios/grouppulse --version

Output is the tool's result text (the upstream body on success). Exit codes: 0 on success, 1 on a runtime error, 2 on a usage error.

MCP server

npx -y @reneturcios/grouppulse mcp

MCP client config:

{
  "mcpServers": {
    "grouppulse": {
      "command": "npx",
      "args": ["-y", "@reneturcios/grouppulse", "mcp"]
    }
  }
}

Docker

docker run -i ghcr.io/cloudcorprecords/grouppulse
docker run --rm ghcr.io/cloudcorprecords/grouppulse get_global_stats '{}'

Development

npm install
npm run build   # tsc -> dist/
npm test        # tsx --test test/*.test.ts (fake fetch only, no network)
npm run dev     # run the CLI from source

License

MIT — Copyright (c) 2026 Rene Turcios

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.