TcGen Agent brings AI assistance into Visual Studio and TwinCAT XAE Shell. It can explain the open project, draft Structured Text, and prepare changes that are applied through the local TwinCAT engineering tools after review.
TwinCAT stays on the engineering workstation. Agile Automation's cloud service handles login, usage limits, model selection, and AI reasoning, while the local bridge reads the project and applies approved updates. TcGen Agent supports TwinCAT 3.1 Build 4024 and 4026+ shell generations.
When you ask for help, TcGen Agent shares the relevant project context with Agile Automation's AI service so the answer is based on your actual TwinCAT solution, not a blank prompt. A local bridge handles project reads and approved writes on the engineering workstation.
TcGen Agent can draft Structured Text and, after approval, update DUTs, functions, interfaces, function blocks, methods, and programs through the standard TwinCAT automation interface.
You choose how project writes are handled: ask every time, approve automatically, or read-only. In ask mode, the generated Structured Text can be reviewed and edited before it is applied.
The local bridge can read source-controlled TwinCAT project files such as .plcproj, .TcPOU, .TcDUT, .TcGVL, and .xti, then give the AI the code it needs to explain the project accurately.
Ask TcGen Agent to produce Structured Text for new blocks, methods, interfaces, data types, and example programs while keeping the generated code visible to the developer.
TcGen Agent can recommend a change, create new TwinCAT objects, or apply updates to an existing solution after approval.
TcGen Agent reads an active TwinCAT project, builds project memory from real objects, explains the architecture, and highlights risks with code-linked evidence.
TcGen Agent designs an object-oriented axis block from project context and Beckhoff motion concepts, applies approved TwinCAT objects, and validates the result.
Baseline usage pool for one engineer
A practical entry point for one engineer evaluating project-aware AI in TwinCAT.
10x Solo usage pool
For teams that want shared access without enterprise procurement overhead.
30x Solo usage pool
For engineering groups using TcGen Agent across active machine software work.
Larger OEMs, self-hosted, privacy review, procurement
For larger organizations that need deployment, privacy, and procurement alignment.
Tell us what you are building, where the current blocker is, and what the target architecture should support.