Follow : Add us as a preferred source on Google.

Key takeaways

• AWS's agents are designed to relieve programmers of having to tie together tasks.
• Amazon claims the agents can run for days without human supervision.
• Amazon is entering a crowded market of vendors promising DevOps agents.

During its re:Invent conference in Las Vegas on Tuesday, Amazon's AWS introduced three agentic AI technologies that the company says will tie together the tasks surrounding mere code writing, including checking code in and out of code libraries and watching for security breaches.

The three agent offerings, dubbed frontier agents, are "a new class of AI agents that are autonomous, scalable, and work for hours or days without intervention," stated AWS in a press release. 

Also: AI agents see explosive growth on AWS Marketplace - over 40x the team's initial targets

The offerings automate tasks in the code repository updates, tasks in DevOps, such as monitoring for application failure, and tasks in cybersecurity, such as analyzing code vulnerabilities.

By performing ancillary tasks in and around code generation, the frontier agents are meant to relieve the burden placed on programmers when they have to manually monitor the input and output of AI coding tools.

With coding tools, "you can find yourself acting as the human 'thread' that holds work together -- rebuilding context when switching tasks, manually coordinating cross-repository changes, and restitching information scattered across tickets, pull requests, and chat threads," stated AWS.

Agents unchaperoned

In its announcement, Amazon emphasized that the frontier agents will work behind the scenes while a programmer is busy with other things.

Learning from its own internal development teams, AWS said, the company realized it was important that "the team could switch from babysitting every small task to directing agents toward broad, goal-driven outcomes."

Also: Enterprises are not prepared for a world of malicious AI agents

AWS also emphasized the importance of "how many agentic tasks they could run at the same time" and that "the longer the agents could operate on their own, the better."

The frontier agents "figure out" how to achieve a goal, "can perform multiple tasks at the same time," and "can operate for hours or days without requiring intervention," stated AWS.

By taking on more tasks without direct oversight, AWS stated, the frontier agents will move "from assisting with individual tasks to completing complex projects autonomously like a member of your team." Amazon did not provide technical details on how the agents operate without oversight or explicit goals. The three frontier agents are:

• Kiro agent for code management can juggle tasks relating to context and repository while a coder works on their principal task, stated AWS. Kiro "maintains persistent context across sessions and continuously learns your pull requests and feedback. It can handle a range of tasks -- from triaging bugs to improving code coverage -- with a single change spanning multiple repositories."
• AWS Security Agent takes an organization's security standards and automatically checks for compliance in design documents and code pull requests. Security Agent will also perform pen testing, a method of trying to compromise an application that's far faster than manual approaches, claimed AWS, and can test multiple applications simultaneously.
• AWS DevOps Agent will automatically perform root cause analysis in the event of an application problem in production. "It learns your resources and their relationships spanning everything from your observability tools, like Amazon CloudWatch, Dynatrace, Datadog, New Relic, and Splunk, to your runbooks, code repositories, and continuous integration and continuous delivery (CI/CD) pipelines," stated AWS.

Also: AI aims to predict and fix developer coding errors before disaster strikes

Developers can access the KIRO agent right now at its dedicated developer site. 

The Security Agent and DevOps Agent are both accessed through the AWS management console. 

A crowded field

With a raft of automations across the development life cycle, AWS is wading into a market crowded with companies that are extending agentic capabilities. Some of these companies have been prominent partners to AWS. 

The DevOps vendors such as Cisco's Splunk, Datadog, and Dynatrace have for years been making the case that their AI-driven automations will speed not only code writing but also testing, debugging, deployment, and monitoring, in an effort to find vulnerabilities before applications go into production.

And code management pure-plays such as GitLab, which competes with Microsoft's GitHub, are rolling out agentic technologies to automatically reconcile code changes. AWS, in fact, has a partnership with GitLab, announced last quarter, to integrate GitLab's agent tools, called Duo Agent, into AWS's generative AI code-writing assistant, Q Developer.

Also: Lost in translation? Amazon Q Developer now speaks more languages

And cybersecurity firms such as Palo Alto Networks have staked out authentication of access to code as one of the responsibilities of a broad enterprise identity and authentication offering, along with automating security alerts for compromised code.

Artificial Intelligence

I'm a Photoshop diehard, but Canva's free tools won me over - and saved me money

AI will cause 'jobs chaos' within the next few years, says Gartner - what that means

I've tried many AI smart glasses in 2025 (including Meta Display) - these are the only ones I'd wear all day

Why AI coding tools like Cursor and Replit are doomed - and what comes next

• I'm a Photoshop diehard, but Canva's free tools won me over - and saved me money
• AI will cause 'jobs chaos' within the next few years, says Gartner - what that means
• I've tried many AI smart glasses in 2025 (including Meta Display) - these are the only ones I'd wear all day
• Why AI coding tools like Cursor and Replit are doomed - and what comes next