Software-Defined IEC 61131-3 Code Generation Agent with Claude Agents SDK
Transitioning industrial automation to software-defined controls requires converting high-level control specs into standard IEC 61131-3 Structured Text. Build a generation and validation agent using Claude Agents SDK and GitHub Copilot patterns to synthesize compile-ready Structured Text code. Achieve 100% compilation success rate on standardized control loop benchmarks.
What you are building
The core problem, expected build, and operating context for this challenge.
Utilize Claude Agents SDK with extended thinking to synthesize and refine IEC 61131-3 Structured Text control logic based on process specifications.
How work is evaluated
Validates generated IEC 61131-3 Structured Text code for compilation accuracy and compliance with functional process specs.
Challenge datasets
Public datasets, plus the private datasets you have access to.
How submissions are scored
These dimensions define what the evaluator checks and which criteria separate a passable run from a strong one.
compiler_pass
Ensures the generated Structured Text compiles without syntax errors.
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
compilation_rate
Percentage of generated specifications successfully compiled. • target: 1 • range: 0-1
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
What you should walk away with
Implement Claude Agents SDK extended thinking capabilities for control logic design
Generate IEC 61131-3 Structured Text (ST) routines compliant with industrial PLC compilers
Verify state machine transitions and safety interlocking constraints using automated check tools
Reference links and supporting material
Natural language control specifications and corresponding validated IEC 61131-3 Structured Text code snippets.
Challenge setup
Validates generated IEC 61131-3 Structured Text code for compilation accuracy and compliance with functional process specs.
Challenge input
JSON containing natural language process specification and I/O map
Claude Agents SDK
Reasoning and agent orchestration
GitHub Copilot
Code pattern integration
Evaluated output
JSON with compiled_status boolean, compiler_errors, and structured_text_code
- Ensures the generated Structured Text compiles without syntax errors.
- Percentage of generated specifications successfully compiled. • target: 1 • range: 0-1
- Benchmark: PLC ST Code Synthesis Benchmark
- Compilation Rate target: 100%
- 1 public reference case
- Python execution harness
- Python sandbox (unavailable on Versalist)
Summary shows the first task, up to three tools, and up to three evidence items. Open the technical recipe to view all configured items.
View technical recipe
Configured tools
- Claude Code · Required
- GitHub Copilot · Optional
- Claude Code · Required
- GitHub Copilot · Optional
- Claude 4.1 Opus · Optional
Evaluation contract
- compiler_pass · Weight 1
- compilation_rate · Weight 1
Metrics
- compilation_rate · Target 1
Benchmarks
- PLC ST Code Synthesis Benchmark
- robotics
Reference cases
- 1 public reference case
Workflow tasks
- Task 1
Input: JSON containing natural language process specification and I/O map
Output: JSON with compiled_status boolean, compiler_errors, and structured_text_code
Execution harness: Python
Evaluation tools
- Claude Agents SDK · framework
- GitHub Copilot · ai-engineering-tooling
Recipe freeze marker
No recipe freeze time is recorded.