すべてのプロジェクト

jevshield

lgy1027/jevshield

Sub-100ms security gate for AI agent tool calls, powered by TypeSafe's Jev (System-1) decision model. Single-request Choice/Noul/Score evaluation, dual-factor blocking matrix, calibrated-confidence routing, fail-closed parsing, zero-config local fallback. LangChain-ready.

Python
スター
5
フォーク
0

掲載理由

Jev/System One で意図の振り分けとツール呼び出しを判断する中立の制御層。編成・再試行・最終返信はアプリに残す。

関連プロジェクト

  • turnstile

    Guardrails for AI agent actions. Deterministic policy, Jev semantic checks, and replayable decisions.

    共通トピック: agent-security, ai-agents, guardrails

  • jevegis

    Guardrails for LLM apps in one API call. Prompt injection, jailbreaks, leaks, unsafe content. Built on TypeSafe Jev. MIT.

    共通トピック: guardrails, llm-security

  • jev-guard

    Vets an LLM agent's tool calls through TypeSafe AI's Jev before they run — allow, block, or hold, failing safe on uncertainty.

    共通トピック: ai-safety, guardrails