Sunday 27 September 2026
Afternoon EditionRSS

ZOTPAPER

News without the noise

ZOTPAPER
AI

AWS open-sources Strands Harness, a general-purpose AI agent for developers

Preconfigured agent builds on the Strands Agents SDK, runs locally or in the cloud and claims a cost edge over rival coding agents

edit
By LineZotpaper
Published
Updated
Read Time3 min
Sources2 outlets
Amazon Web Services has open-sourced Strands Harness, a new general-purpose AI agent designed to give developers a ready-made foundation they can run locally or deploy to the cloud. The release builds on the Strands Agents SDK and positions AWS against rival coding agents such as Claude Code and Codex, with the company claiming a 45% cost advantage.

Amazon Web Services (AWS) has lifted the lid on Strands Harness, an open-source, general-purpose AI agent intended to give developers a preconfigured foundation for building and running AI agents, either locally or in the cloud.

Strands Harness builds on Strands Agents, an open-source Python SDK for building AI agents that AWS debuted in May 2025. The SDK takes what AWS calls a "model-driven approach": developers provide the model, tools and instructions, while the model determines how to tackle the task and when to call those tools. AWS later brought Strands to TypeScript and in February created Strands Labs as a separate home for more experimental projects.

Marc Brooker, vice president and distinguished engineer at AWS, said Strands Harness sits above the existing Strands SDK, giving developers a preconfigured Strands Agent that bundles the tools and supporting machinery an agent needs to operate over longer-running tasks, with AWS supplying its own defaults for how those pieces work together.

"An SDK like the Strands Harness SDK gives you the building blocks, but you still need to decide how to manage context, persist conversations, integrate tools, and guide the agent's behavior," Brooker explained.

Out of the box, Strands Harness provides a working agent with file, shell and web tools, alongside built-in handling for context, memory, persistent sessions, prompt caching and delegation to other agents. Developers can tailor individual agents for different jobs by changing their instructions, choosing which model they use, controlling which tools are available and deciding whether agents can hand work off to one another.

Most of Strands Harness does not depend on AWS infrastructure. The agent loop, tools, context management, session handling and delegation are all included in the open-source release and run on the machine running the agent by default. The exception is model access, which AWS routes through Amazon Bedrock, its managed service for foundation models — a default Brooker described as the only out-of-the-box tie to AWS infrastructure.

AWS says Strands Harness is 45% cheaper than rival tools such as Claude Code and Codex, positioning the open-source agent as a lower-cost foundation for development teams building on its ecosystem.

§

Analysis

Why This Matters

  • Strands Harness lowers the barrier to building AI agents: developers get a working agent out of the box — with file, shell and web tools — rather than assembling context management, memory and tool integration from scratch.
  • It opens a direct open-source challenger to proprietary coding agents like Claude Code and Codex, with AWS claiming a 45% cost advantage.
  • The design signals AWS's strategy of open-sourcing the agent machinery while keeping model access tied by default to its own Bedrock managed service.

Background

AWS introduced Strands Agents in May 2025 as an open-source Python SDK for building AI agents, based on a model-driven approach where the model decides how to tackle tasks and when to invoke tools. The SDK later gained TypeScript support, and in February AWS established Strands Labs as a separate home for experimental projects. Strands Harness is the next step: a preconfigured agent built on that SDK that bundles the supporting machinery for longer-running tasks, with most components running independently of AWS infrastructure.

Key Perspectives

AWS and developers: The open-source release gives developers a ready-made foundation they can run locally or in the cloud, with defaults that can be reconfigured — instructions, model choice, tool availability and inter-agent delegation — making it adaptable to specific jobs. Skeptics: The out-of-the-box setup still routes model access through Amazon Bedrock, tying the default configuration to AWS's managed service, which may limit portability for teams outside the AWS ecosystem. Rival tools and the market: Against established coding agents such as Claude Code and Codex, AWS's cost claims will invite close scrutiny, and adoption may hinge on how Strands Harness compares on real-world performance and flexibility.

What to Watch

  • How easily the Bedrock model-access default can be swapped for other providers in practice.
  • Community traction around Strands Labs and whether the harness gains adoption beyond AWS-centric teams.
  • Further additions from the Strands Labs pipeline — such as new language support or integrations — and independent benchmarks against Claude Code and Codex.

Sources

newspaper

Zotpaper

Articles published under the Zotpaper byline are synthesized from multiple source publications by our AI editor and reviewed by our editorial process. Each story combines reporting from credible outlets to give readers a balanced, comprehensive view.

Editorial ProcessSubscribe