# Amsterdam Open Data Explorer MCP. AI Agent Connect

> Amsterdam Open Data Explorer MCP acts as a direct interface to the City of Amsterdam's DSO Dataplatform. Your AI client can discover over 122 datasets, inspect API specifications, and query live data tables without needing an API key. It turns the city's entire data catalog into a searchable, actionable resource for your agent.

## Overview
- **Category:** data-analytics
- **Price:** Free
- **Endpoint:** https://edge.vinkius.com/vk_preview_tvhA0dZJdMRAGLcaxPco8m1ZxmqgLLr4VFPfp8MM/ai-agent-connect
- **Tags:** amsterdam, open-data, metadata, explorer, catalog, dso-dataplatform, government-public-data

## Description

You can now let your agent navigate Amsterdam's open data ecosystem as if it were a local database. Instead of manually searching through web portals, you use this MCP to let your AI client crawl the DSO Dataplatform. It can find specific datasets by keyword, pull the exact OpenAPI specs for any resource, and fetch actual rows of data using standard filters and pagination. Whether you are looking for specific water management stats or checking the city's data product catalog, this MCP provides the bridge between your agent and the city's live information. It handles the heavy lifting of navigating the platform's structure, so your agent can focus on analyzing the information it finds.

## Tools

### list_datasets
This tool finds all open datasets published by Amsterdam. You can filter the results using specific keywords or tags.

### get_dataset_info
This tool retrieves metadata and OpenAPI specifications for a single dataset. It shows you exactly which resources and operations are available.

### list_data_products
This tool accesses the city's catalog of official data products. You can use keywords or pagination to find what you need.

### browse_dataset
This tool fetches actual data from any dataset resource. You can apply filters, sort results, or use pagination and field projection.

### list_data_teams
This tool identifies the specific city data teams that own the various data products.

## Prompt Examples

**Prompt:** 
```
Which Amsterdam datasets are about water?
```

**Response:** 
```
The list_datasets tool with a 'water' keyword returns matching services, including their descriptions, tags, and API URLs.
```

**Prompt:** 
```
What resources does the bomen dataset expose?
```

**Response:** 
```
Using get_dataset_info for 'bomen' provides the OpenAPI spec and lists available resources like stamgegevens and kapenherplant.
```

**Prompt:** 
```
Fetch the first 10 rows of bag nummeraanduidingen.
```

**Response:** 
```
The browse_dataset tool retrieves the first 10 address rows from the 'bag' dataset using the 'nummeraanduidingen' resource.
```

## Capabilities

### Dataset Discovery
Your agent finds relevant datasets using keywords or tags.

### API Inspection
The agent reads OpenAPI specs to understand how to interact with specific resources.

### Live Data Fetching
Your agent pulls actual records from city tables using filters and pagination.

### Catalog Browsing
The agent identifies data products and the teams responsible for them.

## Use Cases

### Automated Data Discovery
An agent searches for all datasets related to 'transportation' to build a local knowledge base.

### Real-time Data Analysis
A researcher queries the latest water level datasets to check for specific environmental trends.

### API Prototyping
A developer uses the agent to test how different dataset resources respond to specific queries.

### Organizational Mapping
A user identifies which city data teams own specific products for contact or inquiry purposes.

## Benefits

- No API keys are required to access the DSO platform.
- The agent discovers dataset structures automatically via OpenAPI specs.
- Direct table access allows for precise data filtering and sorting.
- The entire city data catalog is accessible through a single connection.

## How It Works

Connecting to this MCP gives your agent immediate access to Amsterdam's data infrastructure.

1. Connect your MCP-compatible client to Vinkius.
2. The agent uses list_datasets to find relevant information.
3. The agent calls get_dataset_info to understand the API structure.
4. The agent uses browse_dataset to pull the specific data rows needed.

## Frequently Asked Questions

**Do I need an API key for Amsterdam open data?**
No, this MCP allows you to explore and query the DSO platform without an API key.

**How many datasets can I access?**
You can access all 122+ datasets published on the Amsterdam DSO Dataplatform.

**Can I filter the data I retrieve?**
Yes, the browse_dataset tool supports filters, sorting, pagination, and field projection.

**Which AI clients can use this MCP?**
You can use this with any MCP-compatible client like Claude, Cursor, Windsurf, or VS Code.

**How does the agent know which API endpoints to use?**
The agent uses the get_dataset_info tool to read the OpenAPI specification for any given dataset.

**Do I need an API key?**
No. The DSO dataset index, the per-dataset OpenAPI specs and the generic table access are public (OPENBAAR). Some individual datasets are marked as requiring a key in their index entry — the dataset info response shows the authentication requirement, so check it before assuming access.

**How do I find the right resource names?**
Call get_dataset_info for the dataset: it fetches the live OpenAPI spec and lists the exposed resources and their operations. Use the resource name (usually the Dutch word) with browse_dataset — e.g. 'stamgegevens' for bomen, 'nummeraanduidingen' for bag.

**What does browse_dataset return?**
A page of the dataset's rows: the items array, current page number and page size, plus a next-link (next) when more pages exist. Filters are passed as comma-separated field=value pairs and only applied when the dataset actually supports that filter field.

**What is the difference between the catalog and the datasets?**
The dataset index (list_datasets) covers the machine-readable REST APIs. The data catalog (list_data_products / list_data_teams) covers the city's curated catalogue of data products and the 26 data teams that produce and own them — useful for mapping a question to the right dataset or owner.