Maintaining and evolving legacy COBOL applications continues to be a challenge for many organizations. Over time, systems grow in complexity, documentation becomes fragmented, and understanding behavior often requires significant manual effort.
Improving visibility with CodeRevive
With CodeRevive for COBOL, we addressed this by creating transparency across the entire system landscape. The January 2026 updates introduced deeper analysis capabilities, including graph-based insights into programs, paragraphs, copybooks, and dependencies, along with integrated JCL analysis for better traceability of batch processes.
At the same time, automated documentation and an enhanced agent made it easier to interact with this knowledge more efficiently.
Users could already work with a built-in assistant to ask questions, explore dependencies, and understand program behavior without navigating the code manually. This made CodeRevive a powerful assistant — with one limitation: the interaction was tied to a dedicated interface.
Accessing this knowledge required switching into the CodeRevive chat assistant.
Introducing the CodeRevive MCP Server
With the introduction of the CodeRevive MCP Server, this limitation is removed.
CodeRevive can now be integrated directly into existing tools and workflows using the Model Context Protocol (MCP).
The knowledge is no longer bound to a single interface, it becomes available where you already work, whether that is in IBM watsonx Code Assistant for Z, IBM Bob, Visual Studio Code, Claude, or other MCP-compatible environments.
Same capabilities, now embedded in your workflow
The way you work with CodeRevive does not change.
You can still:
- analyze program behavior
- understand dependencies
- investigate incidents
- explore system logic
What changes is the context in which you do it.
These capabilities are now accessible directly within your development and analysis environments, without interrupting your workflow.
This integration makes a practical difference in day-to-day work.
Questions that previously required switching tools can now be answered in place, next to your code, during debugging sessions, or within your existing AI assistant.
COBOL knowledge becomes part of the workflow, not a separate step.
Secure by design
Opening access does not mean losing control.
The CodeRevive MCP Server includes secure authentication mechanisms, ensuring that access to system knowledge remains restricted to authorized users and environments.
From standalone solution to integrated workflow
With the introduction of the MCP Server, CodeRevive moves from a standalone solution to an integrated component within a broader toolchain, without changing its core focus: making COBOL systems easier to understand.
At the same time, the MCP Server builds on the existing foundation and opens it up to a wider ecosystem.
The result is a more flexible, more accessible, and more efficient way to work with legacy systems directly within the workflows where that knowledge is needed.
If you’re interested in seeing how this works in your environment, feel free to reach out, we’re happy to show how it works in practice.
