pi-typesafe-jev

A pi extension that exposes TypeSafe (Jev, System One) judgments as five pi tools, so a model can make narrow semantic judgments while your code and your users keep control of thresholds, weights, and actions.

Use cases

  • The file is the question. type , instructions , and criteria come from the file only; the call
  • state is a template. Every { $bind: name } node is replaced by bind.name ; every other node is
  • Every slot must be supplied and every supplied name must be a slot. Both directions fail loudly so a
  • questionFile may be relative (resolved against the current working directory), absolute, or
  • Tool output, details.questionFile , and error messages show the path relative to the working
  • An optional top-level description is echoed in the tool output; any other top-level key is an error.

Source: community catalog. Details and metrics reflect the published community snapshot and are not a performance endorsement. Documentation details extracted from GitHub · the README.

Keep exploring

Similar to pi-typesafe-jev

Browse all
jev-harness (TypeSafeAI) project preview
A custom coding harness for TypeSafe AI's Jev: an LLM proposes, Jev answers narrow questions, code decides, every step leaves a receipt.
jev-enforce project preview
Agent tooling
📏 Claude Code plugin that makes Claude follow your CLAUDE.md: every reply and edit checked by TypeSafe Jev ✅
claude-jev-plugin project preview
Agent tooling
TypeSafe Jev semantic guardrails for Claude Code
heist-one project preview
Games & simulations
Observable browser stealth game: Jev makes typed guard judgments while deterministic code owns the world.
zod-jev project preview
SDKs & clients
Zod validates the shape, Jev validates the meaning: semantic checks on request bodies become calibrated probabilities you threshold in code.
jev-model-tokengate project preview
Benchmarks & research
An OpenAI-compatible proxy that sits between your LLM and your users. It evaluates each sliding window of tokens while the response is still streaming and cuts the stream before a violating token can reach the screen.
nlgrep project preview
Applications
TypeScript CLI that finds code, docs, logs, and text by natural-language conditions using Jev Noul judgments, with probability thresholds, source lines, and cached results.
ai-cli project preview
Integrations
Vercel Labs terminal CLI that can run Jev as the evaluation model for its evaluate command.
jevc project preview
Integrations
Compile agent policy prose into deterministic verdict programs: narrow evidence questions for the model, the verdict computed in code. Install: npm i -g jev-compiler