Overview of the Codesys MCP Toolkit

The Codesys MCP Toolkit bridges the gap between AI assistants that understand the Model Context Protocol (MCP) and the industrial automation environment of CODESYS V3. By exposing a set of MCP resources, tools, and prompts that map directly to CODESYS operations, the server lets developers automate routine project tasks—opening and creating projects, managing Programmable Objects (POUs), editing code, and compiling—all from within an AI‑powered workflow. This eliminates the need for manual interaction with the CODESYS IDE, enabling rapid prototyping and iterative development that is especially valuable in safety‑critical or time‑constrained industrial settings.

The toolkit harnesses the built‑in CODESYS Scripting Engine to execute commands. It offers a clean, declarative interface: tools such as , , and perform the corresponding actions, while resources like expose the hierarchical object model of a project. Developers can query, modify, and persist code snippets or entire modules through simple MCP calls, allowing AI assistants to reason about a project’s state and suggest precise changes. Because the server runs as a Node.js process, it can be launched from any MCP‑enabled client (e.g., Claude Desktop), making it a flexible component in mixed tooling environments.

Key capabilities include:

• Project lifecycle management: Open, create, and save projects with a single command.
• POU manipulation: Generate programs, function blocks, functions, and associated properties or methods; set declaration and implementation code snippets programmatically.
• Compilation control: Trigger a full project compile and receive status updates, enabling continuous integration pipelines that are driven by AI insights.
• Structured data exposure: Retrieve project and POU structures as JSON, giving AI assistants a transparent view of the codebase to support advanced analysis or refactoring suggestions.

Typical use cases span rapid feature addition in embedded controllers, automated code reviews where an AI assistant proposes fixes, and educational settings where students can experiment with PLC programming through natural language commands. In a CI/CD pipeline, an AI assistant could read test results and automatically adjust POU code or add new functions based on detected patterns, all orchestrated through MCP calls to the Codesys server.

Integrating this server into an AI workflow is straightforward: configure the MCP client (e.g., Claude Desktop) to launch the binary with the path to the CODESYS executable and the desired profile. Once connected, the assistant can issue high‑level commands like “create a new function block that reads sensor data” and receive immediate feedback on the project’s state. This tight coupling empowers developers to iterate faster, reduce manual errors, and leverage AI reasoning directly within the industrial software stack.