home / mcp / geekbot mcp server

Geekbot MCP Server

Provides an MCP server to connect Geekbot data (standups, reports, polls, and team members) to LLM clients for natural language interactions.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "geekbot-com-geekbot-mcp": {
      "command": "UV-PATH",
      "args": [
        "tool",
        "run",
        "geekbot-mcp"
      ],
      "env": {
        "GB_API_KEY": "YOUR-API-KEY"
      }
    }
  }
}

Geekbot MCP (Model Context Protocol) server lets you connect LLM clients directly to your Geekbot workspace, enabling natural language interactions with standups, polls, reports, and team members right in your conversations. It acts as a bridge so you can query data, fetch results, and post standup updates without leaving your chat or writing separate API calls.

How to use

Once you have Geekbot MCP running and your LLM client configured, you can perform a variety of actions to manage and retrieve Geekbot data through natural language prompts. You can list standups and polls to discover IDs, fetch specific standup reports or poll results with optional filters, view team members, and post new standup reports. For each action, your LLM client will route your request to the MCP server, which will respond with structured data you can present in your conversation.

How to install

Prerequisites: Python 3.10 or newer and the uv runtime installed on your system. You also need an API key from Geekbot to access your workspace data.

Option 1: Install via Smithery

npx -y @smithery/cli install @geekbot-com/geekbot-mcp --client claude

The remote server will automatically stay up to date with each release.

Option 2: Manual installation

Install Python 3.10+ and uv, then install or upgrade Geekbot MCP using uv. Follow the steps below.

Install Python 3.10+ (examples shown for macOS, Ubuntu/Debian, and Windows).

# macOS
brew install [email protected]

# Ubuntu/Debian
sudo apt update
sudo apt install python3.10

# Windows
# Download and install from python.org

Install uv (the runtime) if you don’t have it yet.

curl -LsSf https://astral.sh/uv/install.sh | sh
```

```powershell
powershell -ExecutionPolicy ByPass -c "irm https://astral.sh/uv/install.ps1 | iex"

Install or upgrade Geekbot MCP with uv.

uv tool install --upgrade geekbot-mcp

On Windows, run the same install command in PowerShell.

Configuration

Connect Geekbot MCP to your LLM client by configuring the client to run the MCP server locally and supply your API key. You will also need the path to your uv executable.

Steps to configure your LLM client (Claude Desktop, Cursor, Windsurf, or another MCP-capable client):

1) Get your Geekbot API Key from Geekbot API/Webhooks settings. 2) Locate the uv executable path on your system. 3) Add the Geekbot MCP server configuration to your LLM client’s MCP settings.

Example MCP configuration for your LLM client desktop (embed the path to your uv executable and your API key):

{
  "mcpServers": {
    "geekbot-mcp": {
      "command": "UV-PATH",
      "args": [
        "tool",
        "run",
        "geekbot-mcp"
      ],
      "env": {
        "GB_API_KEY": "YOUR-API-KEY"
      }
    }
  }
}

Troubleshooting and notes

Review tool usage prompts to ensure you explicitly approve each tool action before execution. This prevents unintended data submissions to Geekbot. You can also request a preview of a report before posting to Geekbot to verify its accuracy.

When fetching data with fetch_reports, limit the date range to avoid large data transfers and potential performance issues. The system will apply sensible limits to retrieved results.

Available tools

list_standups

Lists all standups accessible via your Geekbot API key, including standup IDs, names, channels, times, timezones, questions, participants, owner, confidentiality, and anonymity.

list_polls

Lists all polls accessible via your Geekbot API key, including poll IDs, names, times, timezones, questions, participants, and creator.

fetch_reports

Retrieves standup reports with optional filters for standup, user, and date range. Returns report identifiers, reporters, standup IDs, timestamps, and content.

post_report

Posts a standup report to Geekbot and returns identifiers and content of the submitted report.

list_members

Lists teammates you share standups with in your Geekbot workspace, including member IDs, names, emails, and roles.

fetch_poll_results

Retrieves poll results for a specific poll, with total results and per-question results.

weekly_rollup_report

Generates a comprehensive weekly rollup summarizing team standup responses, highlights, risks, next steps, and upcoming launches.