mcpField
B
Bot Shelf

Bot Shelf

Remote
botshelf.devSearch0Updated 6d ago

About Bot Shelf

Bot Shelf is an MCP connector for searching a community directory of Grok Bots using plain-language job descriptions. Results credit creators and link to add bots in Grok; individual listings, recent additions, categories, and reviewed suggestions are also supported.

Key features

  1. 01Search bots by job description
  2. 02View a bot's full listing
  3. 03List recently added bots
  4. 04View categories and counts
  5. 05Submit bots for review

Use cases

  1. 01Find a Grok Bot for a specific task
  2. 02Inspect a bot listing and its creator
  3. 03Browse recent additions or bot categories
  4. 04Suggest a bot for the directory

MCP support

It is a remote MCP server over Streamable HTTP, exposing tools for search, listing lookup, recent additions, categories, and suggestions.

Configuration

Remote servers need no local install. Pick your client, then copy the config or install in one click.

Endpointhttps://mcp.botshelf.dev/mcp
Copy install command
Claude CodeTerminal
Cursor.cursor/mcp.json
VS Code.vscode/mcp.json
Claude Desktopclaude_desktop_config.json
Windsurf~/.codeium/windsurf/mcp_config.json

Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.

Tools

Details

From the project README

Bot Shelf MCP connector

Search Bot Shelf, an independent directory of 9,000+ community Grok Bots, from any MCP client. Describe a job in plain words and get bots that fit, each with its creator credited and a link that adds it in Grok.

https://mcp.botshelf.dev/mcp

Remote server over streamable HTTP. No account or API key needed. It's free and read-only, apart from suggesting a bot.

Tools

ToolWhat it does
find_botsFinds bots for a job you describe in plain words, best match first, with a link that adds each one in Grok.
get_botOne bot's full listing, from its x.ai/bot share link or its Bot Shelf page.
latestThe bots most recently added to Bot Shelf.
list_categoriesThe categories, with how many bots each has.
suggest_botSuggests a Grok Bot for Bot Shelf. A person reviews every suggestion before it's listed.

Add it to Grok Bot

In Grok Bot's settings, add a custom connector named Bot Shelf with the address https://mcp.botshelf.dev/mcp. Then ask: "Find me a bot on Bot Shelf that chases unpaid invoices."

Full setup steps: botshelf.dev/connector

Add it to other MCP clients

{
  "mcpServers": {
    "bot-shelf": {
      "type": "http",
      "url": "https://mcp.botshelf.dev/mcp"
    }
  }
}

Add it as a skill

For Claude Code, Cursor, Codex and other agents that use skills:

npx skills add cradrrat/bot-shelf-mcp

The skill searches Bot Shelf through the connector, and falls back to plain HTTP when no MCP server is set up.

Add it to Hermes Agent

hermes plugins install cradrrat/bot-shelf-mcp --no-enable
hermes plugins enable bot-shelf

This repository is an Agent Plugins v1 package: plugin.json, mcp.json and skills/bot-shelf/SKILL.md, configuration only, with no code to run.

What it sends where

Every tool call goes to https://mcp.botshelf.dev/mcp, which Bot Shelf runs on Cloudflare. Searches and lookups only read the public directory. suggest_bot sends the link you give it, plus an optional note and name, to a review queue, where it's kept for up to 30 days until a person reviews it. Searches aren't stored. Cloudflare rate-limits calls by IP address, and Bot Shelf doesn't keep those addresses. No account, key or stored credential is involved.

Plain files

Prefer files to a connector? See botshelf.dev/llms.txt.

About

Bot Shelf is made by Nomis (@cradrrat). Bot listings belong to their creators. It isn't affiliated with xAI or X.

The connector's source code isn't published here. This repository is its public listing page, skill and Hermes plugin.

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.