Strands Agents
An open-source SDK for building and running AI agents in Python and TypeScript, providing lifecycle controls, tools, MCP support, multi-agent patterns, memory, tracing and evals in-process.
Strands Agents is an open-source SDK for building and running AI agents in Python and TypeScript, aimed at teams who would otherwise write their own agent loop. It runs in the developer's own process with no hosted control plane, and provides lifecycle controls such as turn limits, token budgets, cancellation and stop reasons, alongside tools, structured output, MCP support, multi-agent patterns, memory and sessions, model portability, streaming, guardrails, tracing and evals.
The monorepo contains the Strands harness (a fully assembled agent available for Python and TypeScript), the underlying Python and TypeScript SDKs covering the agent loop, model providers and tools, a strands CLI for prototyping and chatting with a harness agent from the terminal, and the documentation site.
It offers first-class support for Amazon Bedrock, Anthropic, OpenAI and Gemini plus other providers, traces every agent decision by default, and lets hooks intercept any step to log, validate or redirect it.