TL;DR overview
- SonarQube introduces architecture management features that visualize real component dependencies and flag violations of intended structural rules, making architectural drift visible and actionable.
- Teams can explore dependency graphs at the module and package level, identifying unintended coupling that increases change risk and makes codebases harder to evolve.
- Architecture insights are surfaced alongside quality and security findings, enabling a unified view of code health that includes structural integrity—not just line-level bugs and vulnerabilities.
- This feature helps engineering leaders make data-driven decisions about refactoring priorities by quantifying coupling and dependency complexity across the entire codebase.
For years, Sonar has championed code quality, to make software more maintainable, reliable, and secure. We help millions of developers review and improve their code every day. Today, we are announcing a transformative step forward to help teams manage their software at a higher level, with the addition of architecture capabilities in SonarQube.
A foundation for healthy software
Software architecture is not merely an abstract concept; it is essential in determining how systems function and evolve. Architecture consists of the project structure: how the software is organized into distinct components, the relationships between those components, and the design governing their interactions. Good architecture is modular and transparent. It ensures code changes are localized and have a limited impact on the rest of the application.
Conversely, neglecting software architecture in your project leads directly to the erosion of the coherency of the codebase. The impact of this drift on development is insidious and severe. As architectural drift increases, software evolution becomes harder, and developer productivity stalls. Without tooling to help manage software architecture, teams will fail to stop erosion, will realize this too late, and struggle to fix it. This leads to a situation where changes to the application become so expensive, rewriting it is the only solution. This massive undertaking impedes innovation and slows down time to market.
Why architecture matters even more in the era of AI
While architecture has always been fundamental, the need for high-level clarity is becoming urgent due to the increasing volume and velocity of code production. Codebases are already inherently complex, composed of numerous components and their dependencies. As developers leverage AI coding assistants to solve specific tasks, they risk missing the “big picture” of the overall system, accelerating its erosion.
For this reason, as we move into an era of agent orchestration, humans and agents need to understand the constraints and direction of their architecture. They need to ensure they are delivering on the promises of accelerating software development, not just solving the immediate need. We believe Sonar can help ground these new workflows in sound architectural information, empowering developers and agents to build with confidence.
What architecture capabilities are we adding to SonarQube?
We believe the software industry needs a definitive, simple, and accessible solution for teams to manage their project architecture. With the beta launch of the new architecture capabilities in SonarQube, we are providing a practical approach for teams to understand the actual code structure of their projects, formalize their intended architecture, and manage the gaps.
SonarQube is the only solution in the market that is able to provide these architectural capabilities, at scale, across hundreds of thousands of codebases. They are designed to help you govern software architecture through four key features:
1. Discover
SonarQube produces an accurate picture of your project’s architectural state by reverse-engineering it from the code. It helps teams quickly understand the current structure and relationships of components at all levels. The current architecture serves as the ground truth for developers and provides the necessary context for AI agents.
2. Formalize
You can formally define your target structure for new and existing components, along with their relationships to other components. This intended architecture serves as the blueprint for developers to follow while coding. Configuring the intended architecture can be done incrementally with a minimum upfront time investment. You can start defining high-level components and build toward more detail over time, providing flexibility for different project needs.
3. Prioritize
SonarQube performs automated reviews against your established architectural blueprint, finding gaps even with just a few architecture decisions initially configured. These deviations are automatically detected where the actual code differs from the defined intent. SonarQube also avoids representing the same issue in multiple dimensions. This enables teams to prioritize and tackle the root problem.
4. Fix
Architectural problems are broken down into code level issues that are surfaced as actionable input in the existing developer workflow. These issues show up in the quality gate, avoiding further erosion. This ensures timely resolution before architectural drift compounds. SonarQube tracks remaining gaps and progress in detail.
A first step
In summary, these new architecture capabilities in SonarQube enable developers and AI agents to leverage the current architecture as context to write the right code, and the intended architecture to write the code right. This ensures the software remains modular and maintainable, ready to evolve continuously and cost-effectively.
This is, of course, a beta, so expect it to evolve quickly as we receive feedback. That said, we believe it is a solid foundation that already generates significant value for development teams. It is available on SonarQube Cloud for now (SonarQube Server will come later), supporting Java, JavaScript, Typescript, Python, and C#.
Now that the beta is live, you can request access at any time.
Enjoy!