Direct code deployment accelerates early development when iteration speed matters most. Container-based deployment provides the flexibility and control needed as requirements grow, whether that's larger packages, multi-language support, or custom system dependencies.

Container-based deployment is still there and supports any language, handles packages up to 2GB, and gives teams complete control over dependencies and artifact storage.

It also supports Agent-to-Agent (A2A) and Model Context Protocol (MCP) for building sophisticated multi-agent systems and hosting agentic tools.

This deployment method supports Python 3.10 through 3.13 with packages up to 250MB. It works with all major agent frameworks including LangGraph, Strands, CrewAI, and others.

That’s awesome, thank you for sharing! 👏 One minor note, you can remove AWS from the name, it’s the open source Strands Agents SDK 😊