llm-typesafe 0.1a0
Blog · Watch/Read · (gathered 2026-09-23)
- Simon Willison released
llm-typesafe 0.1a0, a new plugin for hisllmCLI tool, designed to integrate with TypeSafe AI’s Jev model. - The Jev model is highlighted as a “System One” or “Decision Model” LLM, specializing in structured outputs rather than general text generation.
- The
llm-typesafeplugin enables users to ask specific types of structured questions, including “noul” (yes/no with a confidence score), choice (selecting from predefined categories), and scoring (ranking against criteria). - To use it, install with
llm install llm-typesafeand obtain an API key from TypeSafe AI.