Stark is a configurable interpreter designed for programming language evaluation and analysis. It provides a flexible execution environment for running, instrumenting, and comparing code across different language constructs and semantics in a controlled and reproducible way. The system is built to support research into programming language behaviour, including comparative analysis, interpreter-level instrumentation, and automated evaluation workflows. Stark is intended for use in experiments where consistency, traceability, and configurability of execution semantics are important. Rather than targeting a single language implementation, Stark is designed as a modular interpreter infrastructure that can be adapted to different languages or language subsets. This makes it suitable for tasks such as benchmarking language features, testing semantic variations, and evaluating program behaviour under different execution rules. Stark is aimed at researchers, systems programmers, and experimenters working in programming languages, compilers, and AI-assisted code generation analysis.