KoboToolbox MCP Server
Enables MCP clients to query and export KoboToolbox field data using natural language, allowing users to ask questions about submissions and export data without manual logins.
README
KoboToolbox MCP Server
A Model Context Protocol server that lets Claude (or any MCP client) read and export field data directly from KoboToolbox.
Ask in plain language — "How many intakes came in this week that need medical attention?" or "Export this project's data as a CSV" — and the assistant queries the live KoboToolbox API and answers from real records, instead of you logging in and clicking through the dashboard by hand.
As of July 2026, no KoboToolbox MCP server existed in the official MCP registry — this is the first.
<!-- DEMO: replace this line with a GIF or video of the MCP Inspector / Claude Desktop running the tools -->
<!--
-->
Why KoboToolbox
KoboToolbox is the most widely used data-collection toolkit in humanitarian and development work. It's open-source, free to use, and runs the field operations of UNHCR and thousands of NGOs — surveys, needs assessments, monitoring — often in places with poor connectivity and no dedicated IT staff.
That last part is exactly why an MCP server matters here. The organizations running on KoboToolbox are the ones least likely to have engineers who can pull and analyze their own data on demand. Today, answering "what does this month's data actually say?" means a person logging in, filtering submissions by hand, exporting a spreadsheet, and counting. This server collapses that into a question typed in chat — and turns any MCP-capable assistant into someone who can operate the tool on the team's behalf.
There are already thousands of MCP servers for developer tooling and office apps. There were none for the platform that humanitarian responders actually run their work on. This closes that gap.
What it does
Three tools, mapped directly onto the KoboToolbox REST API v2. All read/export — this server never modifies or deletes your data.
| Tool | What it does | Key inputs |
|---|---|---|
list_forms |
Lists every form/project in the connected account — name, UID, deployment status, submission count. | (none) |
get_submissions |
Fetches the submitted records for one form, with optional filtering, sorting, and pagination. | assetUid (required), query, limit, start, sort |
export_form_data |
Kicks off a server-side export, waits for it to finish, and returns a download URL. | assetUid (required), format (csv/xls/geojson/spss_labels), fields |
Requirements
- Node.js 18+
- A KoboToolbox account (free — sign up) on either the global server (
kf.kobotoolbox.org) or the EU server (eu.kobotoolbox.org) - Your account's API token
Getting your API token
Log into KoboToolbox, then either:
- Account Settings → Security → Display your API key, or
- visit
https://<your-server>/token/?format=json
Treat this token like a password — it grants full access to your account's data.
Installation
git clone https://github.com/drannakurasova/kobotoolbox-mcp-server.git
cd kobotoolbox-mcp-server
npm install
npm run build
Configuration
Copy the example env file and fill in your token:
cp .env.example .env
KOBO_API_TOKEN=your_kobotoolbox_api_token_here
KOBO_SERVER=kf.kobotoolbox.org # or eu.kobotoolbox.org
Usage
With Claude Desktop
Add this to your claude_desktop_config.json
(macOS: ~/Library/Application Support/Claude/, Windows: %APPDATA%\Claude\):
{
"mcpServers": {
"kobotoolbox": {
"command": "node",
"args": ["/absolute/path/to/kobotoolbox-mcp-server/dist/index.js"],
"env": {
"KOBO_API_TOKEN": "your_token_here",
"KOBO_SERVER": "kf.kobotoolbox.org"
}
}
}
}
Fully quit and reopen Claude Desktop. The kobotoolbox tools will appear in the tools menu, and you can just ask:
"What forms are in my KoboToolbox account?" "Show me the latest 10 submissions for the intake form." "Export that project's data as a CSV."
With the MCP Inspector
The Inspector is a browser-based test harness — the quickest way to exercise each tool directly (and to record a demo).
npm run inspect
It prints an authenticated URL; open it, click List Tools, select a tool, fill in the inputs, and hit Run Tool.
Development
npm run dev # run from source with tsx (no build step)
npm run build # compile TypeScript to dist/
npm start # run the compiled server
npm run inspect # launch the MCP Inspector against the built server
How it works
src/index.ts— defines the MCP server and registers the three tools, with input validation via Zod. Runs over stdio.src/kobo-client.ts— a thin, typed wrapper around the KoboToolbox REST API v2 (token auth, form listing, submission fetch, and the create-then-poll export flow).
Credentials come only from environment variables — nothing is hardcoded, so the same server works against any KoboToolbox account or form.
Roadmap
Possible extensions:
- Write access (create submissions, edit validation status)
- Per-form schema/question introspection so the assistant understands field meanings
- Support for private/self-hosted KoboToolbox instances
License
MIT
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.