jev-guard
CMaintz/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.
- 星标
- 1
- 派生
- 1
收录依据
jev-guard 在执行前将 Agent 提议的工具调用交给 TypeSafe Jev 检查并返回类型化风险判断。
CMaintz/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.
jev-guard 在执行前将 Agent 提议的工具调用交给 TypeSafe Jev 检查并返回类型化风险判断。
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.
共同主题: ai-safety, guardrails
Guardrails for LLM apps in one API call. Prompt injection, jailbreaks, leaks, unsafe content. Built on TypeSafe Jev. MIT.
共同主题: guardrails
Node.js bindings for TypeSafe's Jev — the System One model. Typed decisions, probabilities and confidence, with no text generation and no parsing.
共同主题: guardrails