Amazon Bedrock AgentCore Runtime introduces interactive shells for terminal access into agent sessions

Amazon Bedrock AgentCore Runtime Updates

Amazon Bedrock AgentCore Runtime now supports interactive shells through a new InvokeAgentRuntimeCommandShell API, providing a persistent, PTY-backed terminal directly into a running agent session over WebSocket. This complements the existing InvokeAgentRuntimeCommand API for one-shot execution, offering developers a full terminal experience inside an isolated microVM with features like colors, tab completion, Ctrl+C, terminal resize, and automatic reconnect on network drop.

This update is particularly beneficial for developers hosting coding agents such as Claude Code, OpenAI Codex, and Amazon Kiro on AgentCore Runtime. They can now authenticate, enter the microVM hosting their coding agent, and interact with it like a local terminal: interact with the agent, inspect files, run ad-hoc commands, or debug the environment state. The shell maintains persistent state across commands within the same session, ensuring environment variables, working directory, and command history behave as expected.

What to do

• Use the new InvokeAgentRuntimeCommandShell API for interactive sessions.
• Authenticate and interact with your coding agent through the terminal.
• Inspect files, run commands, and debug the environment state.
• Reconnect automatically after brief network drops or manually after longer disconnects using the runtime session ID and shell ID.

To get started, use the AgentCore CLI command: agentcore exec --it --runtime . For more information, see Interactive Shells (Terminals) and Shell execution in AgentCore Runtime for a comparison of shell modes.

Source: AWS release notes

If you need further guidance on AWS, our experts are available at AWS@westloop.io. You may also reach us by submitting the Contact Us form.