What's New

AWS Transform agents — which automate migration and modernization of Windows workloads, VMware environments, mainframes, and more — are now accessible directly from popular agentic IDEs including Kiro, Claude, Cursor, and Codex. Developers can invoke Transform capabilities via a Kiro Power, agent plugins, or the new AWS Transform MCP (Model Context Protocol) server, all backed by the same underlying job state. IAM role authentication is also now supported, allowing customers to use existing AWS credentials to provision Transform environments and kick off jobs.

How It Works

• Kiro Power: A native integration available in the Kiro marketplace that embeds AWS Transform agent capabilities directly into the Kiro agentic IDE experience, enabling interactive, conversational-style transformations.
• Agent Plugins: Open-source plugins hosted on GitHub that connect Claude, Cursor, Codex, and other compatible agentic IDEs to the AWS Transform backend, allowing developers to trigger and monitor jobs from within their editor.
• MCP Server: The AWS Transform MCP (Model Context Protocol) server exposes Transform capabilities as a programmatic interface, enabling AI agents and automation pipelines to call Transform operations as structured tool invocations.
• Consistent Job State: All surfaces — IDE, web console, and MCP — operate against the same underlying transformation job, so a job started in an IDE can be monitored in the web console and results retrieved back in the IDE without duplication or state drift.
• IAM Role Authentication: Customers authenticate using standard AWS IAM roles and existing credentials, eliminating the need for separate service-specific credentials and enabling fine-grained access control.
• Supported Workloads: The agents cover multiple modernization domains including Windows server migrations, VMware-to-AWS transitions, and mainframe modernization, drawing on AWS's accumulated migration playbooks and automation.

Why It's Important

• Reduces context switching: Developers no longer need to leave their IDE to initiate or manage complex migration jobs; the entire transformation lifecycle can be managed from a single interface.
• Democratizes migration expertise: AWS Transform encodes decades of migration best practices into automated agents, making expert-level modernization guidance accessible to developers who may not be migration specialists.
• Accelerates enterprise modernization: Embedding transformation capabilities into daily developer tooling lowers the friction to start and sustain large-scale migration projects, potentially compressing timelines significantly.
• Consistent state across surfaces: The shared job model means teams can collaborate — one engineer starts a job in Cursor, a manager reviews progress in the web console — without workflow fragmentation.
• IAM integration reduces operational overhead: Leveraging existing AWS credentials and role-based access control simplifies security governance and removes the need to manage a separate credential store for Transform.
• MCP opens automation pathways: The MCP server enables Transform to be composed into broader agentic workflows and CI/CD pipelines, unlocking programmatic, large-scale migration orchestration.

How It's Different

• Multi-surface vs. single-portal: Unlike traditional migration tools that are confined to a dedicated web console, AWS Transform now operates natively inside the developer's existing IDE and agentic toolchain.
• MCP-first extensibility: The adoption of the Model Context Protocol as an integration layer is forward-looking, positioning Transform to work with any MCP-compatible AI agent or orchestration framework, not just today's named IDEs.
• Unified job state: Competing or earlier approaches often require re-importing or re-specifying jobs when switching tools; here, a single job object is shared across all surfaces simultaneously.
• Opinionated automation backed by AWS IP: The agents are not generic LLM wrappers — they encode AWS-specific migration runbooks, compatibility checks, and remediation patterns built from real-world migration engagements.
• IAM-native security model: Rather than introducing a new identity layer, Transform plugs into the existing AWS IAM ecosystem, which is a meaningful operational advantage for enterprises with mature AWS governance postures.
• Open-source plugin distribution: Publishing plugins on GitHub lowers the barrier to adoption, allows community inspection and contribution, and enables organizations to fork and customize integrations for internal toolchains.

When to Prefer It

• When your team lives in an agentic IDE: If developers are already using Kiro, Cursor, Claude, or Codex as their primary coding environment, the native plugins eliminate the need to context-switch to a separate migration console.
• When automating large-scale migrations programmatically: Use the MCP server when you need to orchestrate Transform jobs as part of a broader automation pipeline, CI/CD workflow, or multi-agent system.
• When migrating Windows, VMware, or mainframe workloads: AWS Transform agents are purpose-built for these specific modernization patterns and will outperform generic scripting or manual approaches for these workload types.
• When IAM governance is a hard requirement: Organizations that mandate IAM-based access control for all AWS services will benefit from the new native IAM role authentication rather than workarounds.
• When teams need to collaborate across roles: The shared job state model is ideal when engineers, architects, and project managers need to interact with the same transformation job from different surfaces (IDE vs. web console).
• When you want to leverage AWS migration best practices without deep expertise: Developers or teams new to cloud migration can rely on the embedded agent intelligence to guide decisions rather than building runbooks from scratch.

Availability

• General Availability: AWS Transform agent integrations (Kiro Power, agent plugins, MCP server) are generally available as of May 14, 2026.
• Plugin Distribution: The agent plugin and MCP server are available as open-source on GitHub; the Kiro Power is available in the Kiro marketplace.
• Supported IDEs: Kiro, Claude (Anthropic), Cursor, and OpenAI Codex are explicitly supported at launch; any MCP-compatible agent environment can integrate via the MCP server.
• Supported Workload Types: Windows server modernization, VMware migration, and mainframe modernization are confirmed; additional Transform agent types may vary by region and service tier.
• Authentication: IAM role authentication is now supported; customers use existing AWS credentials to create Transform environments, workspaces, and jobs.
• Pricing: Specific pricing details are not disclosed in the announcement; customers should refer to the AWS Transform pricing page at aws.amazon.com/transform for current rates.
• Regional Availability: Specific supported AWS regions are not enumerated in the announcement; consult the AWS regional services list or the Transform documentation for region-by-region availability.