MCP server · Data + databases
DuckDB / MotherDuck MCP
Query local DuckDB files or MotherDuck through MCP, with read-only defaults and optional write-enabled SQL.
Quick verdict
It pairs DuckDB's low-friction embedded analytics with a small official tool set, needs no cloud credentials for local files, and can reach MotherDuck later without changing servers.
- MaintainerMotherDuck (official)
- Security scopeLocal execution
- Transportsstdio, streamable-http
- Install channelpypi · mcp-server-motherduck
- EvidenceSource Validated
- LicenseMIT · Open source
MotherDuck's official server for browsing catalogs and running DuckDB SQL across in-memory databases, local files, S3 objects and MotherDuck itself. One canonical open-source implementation covers both names; the configuration published here opens a single local DuckDB file in the upstream's default read-only mode. Write access and database switching exist but must be turned on explicitly.
Install DuckDB / MotherDuck MCP
uvx mcp-server-motherduck --db-path
{
"mcpServers": {
"duckdb": {
"command": "uvx",
"args": ["mcp-server-motherduck", "--db-path", "<ABSOLUTE_PATH_TO_DATABASE>"]
}
}
}Runs with uvx (uv); the CLI names its HTTP transport `http`. The published configuration deliberately omits --read-write and --allow-switch-databases. A self-hosted HTTP container is also documented upstream.
Credentials
| Name | Required | What it is for |
|---|---|---|
motherduck_token | Optional | MotherDuck token, needed only for md: connections. The local DuckDB configuration published here does not use it. |
AWS_ACCESS_KEY_ID | Optional | Optional AWS access key for S3-backed databases. |
AWS_SECRET_ACCESS_KEY | Optional | Optional AWS secret key for S3-backed databases. |
AWS_SESSION_TOKEN | Optional | Optional session token for temporary AWS credentials. |
Verified configurations
Not run by us yet. This profile is Source Validated: the install channel, config entry, credentials and license were checked against the source and upstream metadata, but no configuration has been installed, discovered and executed in a recorded environment. When one is, it appears here with the eight scope dimensions below.
Best for
Local analytics and data-engineering work where an agent needs SQL over a controlled DuckDB file without standing up a database server.
Why use it
It pairs DuckDB's low-friction embedded analytics with a small official tool set, needs no cloud credentials for local files, and can reach MotherDuck later without changing servers.
Tradeoffs
This is local SQL execution, not a read-only API, and upstream states plainly that read-only mode is not a complete sandbox — it does not prevent local filesystem access or DuckDB setting changes. That is why the security scope here is local-execution rather than read-only. Enabling --read-write or --allow-switch-databases widens the blast radius considerably, and S3 or MotherDuck modes add credentials and cost. Point it at one controlled database file.
Upstream state
pypi package published · repository active · in the official MCP Registry (active). Lifecycle is derived from a committed upstream snapshot at build time, never authored by hand.
Official MCP Registry name: io.github.motherduckdb/mcp-server-motherduck
Editorial review
Grade B — Source Validated. Re-review is due by 2027-02-15. Config entries, credential names and package status can change upstream; verify against the source repository before granting an agent access.
Sources
Related MCP servers
- Exa MCPSearch the web and fetch clean page content through Exa, with optional advanced search and multi-step research tools.
- Hugging Face MCPSearch Hugging Face models, datasets, Spaces, papers and docs through MCP, with optional Hub and compute tools.
- Redis MCPLet agents read, write, search and manage Redis data structures through Redis's official MCP server.