The AWS API MCP Server acts as a bridge between AI assistants and AWS services, using AWS CLI commands to create, update and manage resources. It helps with command selection and exposes the latest AWS API features. The README notes it is superseded by the official AWS MCP server.
What it does
Where AWS API MCP Server runs and what it adds, from its documentation
Executes AWS CLI commands through natural language
Validates commands before running them
Supports all AWS services via the CLI
Optional read-only mode and mutation consent prompts
Restricts local file access to a working directory by default
When to use it
Good fit
General AWS interactions and exploring resources
Managing infrastructure from an MCP-enabled assistant
Not for
New setups where the README points to the official AWS MCP server as replacement
Sandboxed execution: the README states there is no sandboxing
How to install AWS API MCP Server
You need: AWS account with credentials; Python 3.10 or newer; uv optional
Configure AWS credentials (AWS_API_MCP_PROFILE_NAME or standard boto3 methods)
Ensure Python 3.10+ (and optionally uv) is installed
Add the uvx config to your MCP client
Optionally set READ_ONLY via READ_OPERATIONS_ONLY=true or REQUIRE_MUTATION_CONSENT=true