# Prerequisites for Using the AWS Agent Toolkit: Complete Setup Guide

> Learn the prerequisites for using the AWS Agent Toolkit. Ensure you install uv and configure AWS credentials for seamless setup. Get started today.

- Repository: [Amazon Web Services/agent-toolkit-for-aws](https://github.com/aws/agent-toolkit-for-aws)
- Tags: getting-started
- Published: 2026-06-28

---

**To use the AWS Agent Toolkit, you must install the `uv` Python package manager and configure valid AWS credentials on your local machine.**

The AWS Agent Toolkit is a collection of plugins, skills, and a managed AWS MCP Server that enables AI coding agents to interact with AWS services. Before you can install the `aws-core` plugin or invoke AWS services through your agent, you must satisfy two essential prerequisites documented in the `aws/agent-toolkit-for-aws` repository.

## Essential Prerequisites for the AWS Agent Toolkit

The toolkit requires two foundational components to function correctly: the `uv` runtime for Python execution and valid AWS authentication for API access.

### Install the `uv` Package Manager

The toolkit relies on **`uv`** as its Python package manager and runtime. According to the repository's top-level [[`README.md`](https://github.com/aws/agent-toolkit-for-aws/blob/main/README.md)](https://github.com/aws/agent-toolkit-for-aws/blob/main/README.md), `uv` is mandatory for running the toolkit's scripts and MCP Server client.

To install `uv` on Linux or macOS:

```bash
curl -LsSf https://astral.sh/uv/install.sh | sh

```

Verify the installation was successful:

```bash
uv --version

```

### Configure AWS Credentials

You need a valid AWS account with credentials configured locally so the MCP Server can authenticate API calls. The [[`README.md`](https://github.com/aws/agent-toolkit-for-aws/blob/main/README.md)](https://github.com/aws/agent-toolkit-for-aws/blob/main/README.md) explicitly states that an AWS account with credentials configured locally is required for API calls and script execution.

Configure your credentials using the AWS CLI:

```bash
aws configure

```

This prompts for your Access Key ID, Secret Access Key, default region, and output format. Alternatively, you can use environment variables or IAM roles if running on AWS infrastructure.

## Installing the Core Plugin

Once `uv` and AWS credentials are in place, you can install the `aws-core` plugin. The [[`plugins/aws-core/README.md`](https://github.com/aws/agent-toolkit-for-aws/blob/main/plugins/aws-core/README.md)](https://github.com/aws/agent-toolkit-for-aws/blob/main/plugins/aws-core/README.md) assumes these prerequisites are already satisfied before proceeding.

For **Claude Code**:

```bash
/plugin install aws-core@claude-plugins-official
/plugin reload

```

For **terminal-based agents** like Codex:

```bash
codex plugin marketplace add aws/agent-toolkit-for-aws

```

## How Prerequisites Enable Core Functionality

These prerequisites directly support the toolkit's architecture. The [[`plugins/aws-core/skills/signing-in-to-aws/SKILL.md`](https://github.com/aws/agent-toolkit-for-aws/blob/main/plugins/aws-core/skills/signing-in-to-aws/SKILL.md)](https://github.com/aws/agent-toolkit-for-aws/blob/main/plugins/aws-core/skills/signing-in-to-aws/SKILL.md) file describes the skill agents use to authenticate with AWS, which depends on the pre-configured credentials being present. Similarly, the [`rules/`](https://github.com/aws/agent-toolkit-for-aws/tree/main/rules) directory contains optional rule files that require a working MCP Server, which in turn depends on `uv` and AWS authentication.

## Summary

- **Install `uv`** to provide the Python runtime and package management required by the toolkit's scripts and MCP Server client.
- **Configure AWS credentials** locally to enable the MCP Server to authenticate API calls and execute AWS service operations.
- The `aws-core` plugin and all downstream skills depend on these two prerequisites being satisfied first.
- Optional components in the `rules/` directory and agent-specific connectors build upon this foundation.

## Frequently Asked Questions

### What is `uv` and why is it required by the AWS Agent Toolkit?

`uv` is a fast Python package manager and runtime developed by Astral. The AWS Agent Toolkit uses `uv` to execute scripts and manage the MCP Server client, making it a mandatory prerequisite according to the repository's [`README.md`](https://github.com/aws/agent-toolkit-for-aws/blob/main/README.md).

### Can I use AWS SSO or IAM roles instead of access keys?

Yes. While the examples show `aws configure` with access keys, the toolkit only requires that valid AWS credentials are present in your environment. This can include AWS SSO profiles, IAM roles on EC2 instances, or credentials from AWS IAM Identity Center, as long as they are accessible to the MCP Server process.

### Which AI agents are compatible with this toolkit?

The toolkit supports multiple agents including Claude Code, Codex, Cursor, and Kiro. Each agent has specific UI steps for plugin installation, but all require the same two prerequisites: `uv` installed and AWS credentials configured locally.

### Where can I find additional configuration options?

Detailed configuration instructions are documented in the [`plugins/aws-core/README.md`](https://github.com/aws/agent-toolkit-for-aws/blob/main/plugins/aws-core/README.md) file. For authentication-specific guidance, refer to [`plugins/aws-core/skills/signing-in-to-aws/SKILL.md`](https://github.com/aws/agent-toolkit-for-aws/blob/main/plugins/aws-core/skills/signing-in-to-aws/SKILL.md). The top-level [`README.md`](https://github.com/aws/agent-toolkit-for-aws/blob/main/README.md) in the repository provides the complete quick-start guide.