AWS has officially released Aura MySQL MCP Server! It now allows querying using natural language.
This is Yuta Kikai (@fat47) from the Service Reliability Group (SRG) of the Media Division.
#SRGThe Service Reliability Group primarily provides comprehensive support for the infrastructure surrounding our media services, focusing on improving existing services, launching new ones, and contributing to open-source software (OSS).
This article is a review of the Aurora MySQL MCP Server released by AWS.
I hope this is of some help.
awslabs AWS MCP ServersAmazon Aurora MySQL MCP ServerCreating a verification tableLet's try setting up MCP in Cursor and executing queries using natural language.The system automatically collects data on slowdowns and suggests improvement plans.About risksIn conclusion
awslabs AWS MCP Servers
The official AWS MCP server cluster can be found here.
When it was first released, there were only a few types, but when I checked recently, the variety had increased dramatically.
Even just picking out some of the most interesting ones from recent releases, there are this many.
Amazon Aurora MySQL MCP Server
I'm particularly interested inAmazon Aurora MySQL MCP ServerI tried verifying it.
The documentation containing the necessary information for implementation can be found here.
I will quote the features from the above document.
Features Natural language to MySQL SQL query Converting human-readable questions and commands into structured MySQL-compatible SQL queries and executing them against the configured Aurora MySQL database.
function
Natural Language to MySQL SQL Query Conversion
Converts human-readable questions and commands into structured, MySQL-compatible SQL queries and executes them against the configured Aurora MySQL database.
RDS Data APIRestrictions on the use of the RDS Data API
- Not available for burstable instance classes (t3, t4)
- Aurora MySQL version 3.07 or later is required.
It is important to note that there is a certain point.
To enable the RDS Data API on an existing cluster that meets the above requirements,
Change the settings that appear when you scroll all the way down to the bottom of the "Connection and Security" tab in an existing cluster.
Click to enable the RDS Data API.
Creating a verification table
Connect to Aurora MySQL and create two tables for testing.
Insert several dozen appropriate data entries into each table.
Sample INSERT statement
Let's try setting up MCP in Cursor and executing queries using natural language.
This section provides an example of using Cursor in MCP configuration.
Create a .cursor/mcp.json file in your working directory.
You will give instructions in the chat box next to the cursor.
You will then be prompted to grant permission to start Docker as defined in mcp.json, so press "Run".
テーブル一覧を取得してください
SHOW TABLE;
Currently, a semicolon (;) at the end of the suggested query, as described above, will result in an error.
This appears to be an error in the mysql-mcp-server implementation.
; ;
It is also possible to check the structure of a specific table.
The system automatically collects data on slowdowns and suggests improvement plans.
We will automatically collect slow logs and suggest adding an index as an improvement suggestion.
First, it performed the aggregation using mysqldumpslow as instructed.
From here, it automatically uses MCP to check the actual status of the table.
*The sample query only inserted a few dozen records, but to make it a slow query, millions of records have been added.
The analysis results were displayed as follows.
They also provide information on recommended indexes.
Finally, they also include some points to note regarding the addition of an index.
About risks
The ability to execute queries in natural language is extremely convenient. In this case, the MCP server is running in read-only mode, and only read queries are allowed, so major accidents are unlikely to occur.
However, the MCP server needs to be configured for each user's environment, and it may sometimes be started without the read-only restriction.
One possible solution would be to restrict the database user information used by the MCP server to only those with appropriate permissions.
Furthermore, although we only used it in a testing environment this time, when using it in a production-like environment, it is advisable to appropriately restrict the permissions of the IAM user used by the MCP server.
In conclusion
I've come to realize that Amazon Aurora MySQL MCP Server alone seems incredibly useful.
This time, I manually provided slow query logs for analysis, but I'm excited about the possibility of combining it with Amazon Cloudwatch Logs MCP Server to perform everything from log retrieval to query analysis in natural language!
I encourage you all to try out the various official AWS Labs MCP servers.
SRG is looking for new team members.
If you are interested, please contact us here.