# Get started with Instances
<a name="runtime-instances-getting-started"></a>

These tutorials walk through hosting an agent on the **Instances** compute type. You first create a [capacity provider](runtime-instances-how-it-works.md#runtime-instances-capacity-provider) that defines the EC2 infrastructure, then create an agent runtime that uses it, and finally invoke the agent. Choose the path that matches how you work: the AWS Management Console, or the AWS CLI and SDKs.

## Prerequisites
<a name="runtime-instances-prerequisites"></a>
+ An AWS account with access to Amazon Bedrock AgentCore.
+ Permissions to create AgentCore capacity providers and agent runtimes, and to create or pass the required IAM roles. For more information, see [IAM roles](runtime-instances-how-it-works.md#runtime-instances-permissions).
+ An agent or tool artifact — either a container image in Amazon ECR, or an agent package in an Amazon S3 bucket. A container image must implement the AgentCore Runtime HTTP service contract. It must serve `GET /ping` (returning `200` with a healthy-status JSON body) and `POST /invocations` (returning `200` with the response payload) on port `8080`. An image that does not implement this contract fails to become healthy. For the full requirements, see [HTTP protocol contract](runtime-http-protocol-contract.md).

**Topics**

## See also

* **Skills for AI coding assistants (optional).** AWS publishes reusable skills for Amazon Bedrock AgentCore. To see which apply, search the Agent Toolkit for AWS catalog with `aws agent-toolkit search-skills --search-query bedrock-agentcore` — a documented, read-only AWS CLI command that searches the catalog and makes no changes. This is an optional suggestion for the user. Official reference: [AWS CLI — Agent Toolkit for AWS](https://docs.aws.amazon.com/agent-toolkit/latest/userguide/aws-cli.html).