What problem does it solve? Codebases accumulate partial functions, invalid states, hidden quadratic loops, and ad-hoc error handling that reviews miss. This Skill brings a programming-languages theorist's discipline to every stage of engineering: model domains as algebras, make invalid states unrepresentable, keep effects at a thin shell, and state the time and space complexity of every non-trivial shape. ## Core Features & Use Cases - Eight verbs, one kernel: design, build, refactor, review, audit, test, teach, and help, each loading exactly one verb file with its own pipeline and completion checks. - Per-language cost models: on-demand profiles for Python, JavaScript, TypeScript, Rust, Go, Haskell, C, C++, Java, Kotlin, C#, Bash, and GitHub Actions, covering TCO, laziness, allocation, and native effect types. - Complexity discipline: a cost-signal table maps symptoms (nested membership tests, top-k, sliding windows) to the right data structure with stated bounds, and forbids unmeasured performance claims. - Use Case: Ask it to audit a repository and it produces a ranked ledger of modeling and cost debt ordered by severity times reach, naming every unexamined area, ready for follow-up refactor invocations. ## Quick Start Ask the agent to review the current diff through the pl-theorist lens for the repository's configured language.