Skip to main content
Glama

BeeBoo MCP Server

Model Context Protocol (MCP) server for BeeBoo — Human-in-the-Loop Infrastructure for AI Agents.

This server enables AI agents like Claude, Cursor, and Windsurf to natively interact with BeeBoo's capabilities:

  • Knowledge Base — Search, add, and list knowledge entries

  • Approvals — Request and check human approval status

  • Work Requests — Create and track work requests

Quick Start

1. Get your API Key

Get your BeeBoo API key from beeboo.ai/settings/api-keys.

Your key will look like: bb_sk_xxxxxxxxxxxx

2. Install & Configure

Choose your AI tool:

Claude Desktop

Add to your claude_desktop_config.json:

macOS: ~/Library/Application Support/Claude/claude_desktop_config.json Windows: %APPDATA%\Claude\claude_desktop_config.json

{
  "mcpServers": {
    "beeboo": {
      "command": "npx",
      "args": ["-y", "@beeboo/mcp-server"],
      "env": {
        "BEEBOO_API_KEY": "bb_sk_your_key_here"
      }
    }
  }
}

Then restart Claude Desktop.

Cursor

Add to your Cursor settings (~/.cursor/mcp.json or via Settings > MCP):

{
  "mcpServers": {
    "beeboo": {
      "command": "npx",
      "args": ["-y", "@beeboo/mcp-server"],
      "env": {
        "BEEBOO_API_KEY": "bb_sk_your_key_here"
      }
    }
  }
}

Windsurf

Add to your Windsurf MCP configuration:

{
  "mcpServers": {
    "beeboo": {
      "command": "npx",
      "args": ["-y", "@beeboo/mcp-server"],
      "env": {
        "BEEBOO_API_KEY": "bb_sk_your_key_here"
      }
    }
  }
}

Alternative: Local Install

npm install -g @beeboo/mcp-server

Then use beeboo-mcp-server instead of npx @beeboo/mcp-server.

Related MCP server: Finizi B4B MCP Server

Available Tools

Tool

Description

beeboo_knowledge_search

Search the knowledge base using semantic search

beeboo_knowledge_add

Add a new entry to the knowledge base

beeboo_knowledge_list

List all knowledge base entries

beeboo_approval_request

Request human approval for an action

beeboo_approval_check

Check status of an approval request

beeboo_approvals_list

List all approval requests (with optional filter)

beeboo_request_create

Create a work request for the team

beeboo_requests_list

List all work requests (with optional filter)

Usage Examples

Once configured, you can ask your AI assistant:

Knowledge Base:

  • "Search the knowledge base for deployment procedures"

  • "Add to the knowledge base: our AWS account ID is 123456789"

  • "List all knowledge entries"

Approvals:

  • "I need approval to delete the staging database"

  • "Check if approval abc123 has been approved"

  • "Show me all pending approvals"

Work Requests:

  • "Create a high-priority request to update the SSL certificate"

  • "List all open work requests"

Environment Variables

Variable

Required

Default

Description

BEEBOO_API_KEY

Yes

—

Your BeeBoo API key

BEEBOO_API_URL

No

https://beeboo-api-625726065149.us-central1.run.app

API endpoint

Testing

Test the server locally:

# List available tools
echo '{"jsonrpc":"2.0","id":1,"method":"tools/list"}' | BEEBOO_API_KEY=your_key node index.js

# Test a tool call
echo '{"jsonrpc":"2.0","id":2,"method":"tools/call","params":{"name":"beeboo_knowledge_list","arguments":{}}}' | BEEBOO_API_KEY=your_key node index.js

Troubleshooting

"BEEBOO_API_KEY environment variable is required"

Make sure you've set the BEEBOO_API_KEY in your MCP configuration.

Server not appearing in tools list

  1. Restart your AI tool (Claude Desktop, Cursor, etc.)

  2. Check the configuration file path is correct

  3. Verify the JSON syntax is valid

API errors

  1. Check your API key is valid

  2. Ensure you have network connectivity

  3. Check the BeeBoo status at status.beeboo.ai

Development

# Clone the repo
git clone https://github.com/beeboo-ai/beeboo.git
cd beeboo/mcp-server

# Install dependencies
npm install

# Run locally
BEEBOO_API_KEY=your_key npm start

License

MIT

Available Tools

8 tools
beeboo_approval_checkC

Check the status of an approval request

Inputs: id (required).

beeboo_approval_requestA

Request human approval for an action. Use this when you need explicit permission before proceeding with a potentially impactful operation.

Inputs: title (required), description (required).

beeboo_approvals_listB

List all approval requests with optional status filter

Inputs: status.

beeboo_knowledge_addC

Add a new entry to the BeeBoo knowledge base

Inputs: title (required), content (required), tags.

beeboo_knowledge_listB

List all knowledge base entries

Inputs: no named parameters.

beeboo_request_createB

Create a work request for the team. Use this to queue up tasks that need human attention or execution.

Inputs: title (required), description, priority.

beeboo_requests_listC

List all work requests with optional status filter

Inputs: status.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    A3.6/5.0

    Scored across 8 tools

    Disambiguation5/5

    Every tool has a clearly distinct purpose with no ambiguity. The tools are cleanly separated into three functional groups: approvals (check, request, list), knowledge base (add, list, search), and work requests (create, list). There is no overlap in functionality between these groups or within each group.

    Naming Consistency5/5

    All tools follow a perfect and consistent naming pattern: 'beeboo_' prefix followed by resource name (approval/knowledge/request) and action verb (check/request/list/add/search/create). The pattern is uniformly applied across all 8 tools with no deviations in style or structure.

    Tool Count5/5

    With 8 tools, the server is well-scoped for its purpose of handling approvals, knowledge management, and work requests. Each tool earns its place by providing distinct functionality within these three domains, neither feeling too sparse nor overloaded.

    Completeness4/5

    The tool surface provides excellent coverage for the three domains, with CRUD-like operations for each: approvals (request, check, list), knowledge (add, list, search), and work requests (create, list). Minor gaps include no update/delete operations for knowledge entries or work requests, but agents can work around this limitation.

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      B
      quality
      F
      maintenance
      Connects AI assistants (Claude, Cursor, etc.) directly to the HiveFlow automation platform, allowing them to create, manage, and execute automation flows through natural language commands.
      9
      4 npm
      4
      MIT
    • A
      license
      Not graded
      quality
      Not graded
      maintenance
      Enables AI assistants to interact with the Finizi B4B platform through 15 comprehensive tools for managing business entities, invoices, vendors, and products. Features secure JWT authentication, automatic retries, and full business-to-business workflow integration through natural language commands.
      MIT
    • F
      license
      Not graded
      quality
      D
      maintenance
      Enables AI assistants to interact with the Beep platform to manage bounties, assets, and payments. It supports operations for starting and stopping streaming sessions, verifying transactions, and issuing payments on the Beep network.
      -
    • A
      license
      A
      quality
      D
      maintenance
      Human-in-the-loop approval gate for AI agents. Your agent calls submit_approval before any irreversible action; a human reviews on a branded page; a signed webhook fires back with the decision.
      1
      1
      MIT