Search and Explore AWS Open Data
MCP server for searching and exploring the AWS Open Data Registry - fuzzy dataset search by name, description, or tags, with cached lookups.
Why it matters
Access and query the AWS Open Data Registry to discover and retrieve information about available datasets. This asset facilitates efficient data exploration through fuzzy searching and caching.
Outcomes
What it gets done
Search datasets by name, description, or tags using fuzzy matching.
Retrieve detailed information for specific datasets.
Leverage automatic caching for faster query responses.
Explore datasets from the AWS Open Data Registry.
Install
Add it to your toolbox
Run in your project directory:
curl -fsSL https://spark.entire.vc/get/vb-aws-open-data | bash Capabilities
Tools your agent gets
Search datasets by name, description, or tags with fuzzy search.
Get detailed information about a specific dataset by its file name.
Overview
AWS Open Data MCP Server
An MCP server for searching and exploring the AWS Open Data Registry, with fuzzy dataset search across names, descriptions, and tags, and detailed per-dataset lookups, backed by an automatically cached copy of the registry. Use it when an AI assistant needs to find or inspect a public dataset from the AWS Open Data Registry, such as satellite or climate data, by name, topic, or tag.
What it does
An MCP server that lets an AI assistant search and explore the AWS Open Data Registry, the catalog of publicly available datasets hosted on AWS, without browsing the registry's GitHub repository by hand. It maintains an automatically cached copy of the registry for fast lookups, and exposes fuzzy search across dataset names, descriptions, and tags alongside detailed per-dataset retrieval.
When to use - and when NOT to
Use it when an AI assistant needs to find a public dataset relevant to a topic, such as searching for satellite imagery data, or needs the full details of a dataset already identified by its registry ID. It is scoped specifically to the AWS Open Data Registry's catalog metadata: it helps find and describe a dataset, but does not itself download data, query the dataset's contents, or provision AWS resources to access it. Choosing the detail level appropriately matters for token budget - nameOnly or minimal for a broad search, full only once a specific dataset is the actual target.
Capabilities
search_datasets matches a query against dataset names, descriptions, and tags, with an optional limit (default 25) and a detail level of nameOnly, minimal (name and description, the default), or full (complete dataset info); an empty query returns all datasets. get_dataset takes a dataset's registry ID, without the .yaml extension, for example sentinel-1, and returns its full details.
How to install
npm install
npm start
The server runs at http://localhost:3000/mcp over streamable HTTP transport. Add it to Claude Code with:
claude mcp add aws-open-data --transport http http://localhost:3000/mcp
Restart Claude Code afterward to pick up the new MCP server.
Who it's for
Developers and researchers who want an AI assistant to discover and describe publicly available AWS-hosted datasets, such as satellite imagery or climate data, by name, topic, or tag instead of manually searching the registry.
Source README
AWS Open Data Registry MCP Server
An MCP server that allows searching and exploring datasets from the AWS Open Data Registry.
Features
- search_datasets: Search for datasets by name, description, or tags with fuzzy matching
- get_dataset: Get detailed information about a specific dataset
- Automatic caching of the AWS Open Data Registry for fast lookups
Installation
npm install
Usage
npm start
The server will be available at http://localhost:3000/mcp using the streamable HTTP transport. You can install this in Claude Code with: claude mcp add aws-open-data --transport http http://localhost:3000/mcp. You'll then need to restart Claude Code to pick up the new MCP.
Tools
search_datasets
Search for datasets matching a query. If no query is provided, returns all datasets.
Arguments:
query(string, optional): Search query to match against dataset names, descriptions, and tags. Empty query returns all datasetslimit(number, optional): Maximum number of results to return (default: 25)detail(enum, optional): Detail level - 'nameOnly' (just names), 'minimal' (name and description), or 'full' (complete dataset info). Default: 'minimal'
Example:
{
"query": "satellite",
"limit": 5,
"detail": "minimal"
}
get_dataset
Get detailed information about a specific dataset.
Arguments:
id(string, required): The ID of the dataset (without .yaml extension)
Example:
{
"id": "sentinel-1"
}
FAQ
Common questions
Discussion
Questions & comments · 0
Sign In Sign in to leave a comment.