Guardrails & safety — page 2

DSH plugin: register TypeSafe Jev (System One decision model) as an agent tool — jev_decide returns calibrated probabilities (noul/choice/score) for routing/triage/guardrail judgments, no text generation. 把 TypeSafe Jev 决策模型注册为 DSH agent 工具

Automated database migration safety reviewer powered by TypeSafe AI (Jev System One model)

Automode for DeepSeek Harness: a fourth permission preset — full access with an LLM classifier as the only gate before every tool call.

Self-hosted MCP server for Kalshi prediction markets. Native RSA-PSS auth, token-bucket rate limiting, demo/prod safety controls. Designed to be forked and deployed.

A decision layer for agent control flow — yes/no, choice and rubric questions answered from one token's probabilities, measured against labelled sets — with a Claude Code permission hook and a trainable local judge

Experimental native Codex Desktop and CLI model routing with Jev and a configurable allowlist

Checks whether a coding agent completion claim is actually true: deterministic evidence gathered in code, one batched Jev assessment, and a pure policy. DSH plugin.

TypeSafe Jev (System One decision model) as a decision layer for DeepSeek Harness: typed decisions, confidence-gated routing, a cost meter, and an automatic prompt-injection guard over untrusted tool results.

A UiPath coded agent with a custom PII detection guardrail on the LLM boundary, built on the TypeSafe Jev model as a LangChain awrap_model_call middleware.

Pi extension that uses Jev task classification (via OpenRouter) to route each run to explicitly configured models with fail-open policy. Public Preview.

Alert triage as a UNIX filter: JSONL security alerts in, typed decisions out. Runs on TypeSafe Jev or a local model; policy routing stays in code.

Comment-moderation playground: paste a comment, Jev decides what to do with it.

Jev-augmented Playwright MCP proxy — page-state triage, prompt-injection shielding, goal-based snapshot pruning, risky-action gating. Drop-in wrapper around @playwright/mcp for any coding agent.

Typed verdicts on pull requests with TypeSafe System One (Jev): one parallel call, policy in code, usable as a GitHub Action

Typed, policy-driven decision workflows on top of TypeSafe AI Jev: confidence routing, fallbacks, evaluation, and RAG patterns for TypeScript apps.

OpenCode v2 plugin that sends permission requests resolved to "ask" to a model of your choice for review. Fails closed on every error, never overrides a configured deny, and writes every decision to an audit log.

AI Harness - JIT policy routing, MCP/Skill capability acquisition, shared machine runtime registry

Decide what an AI system should remember: policy-driven memory gating with typed judgments.

Unofficial Jev-powered Zig terminal agent with deterministic policy gates and encrypted audit trails.

Probabilistic pull-request risk triage using Jev and deterministic review policy.

MCP server that lets Claude Code, Codex & AI agents drive a real Chrome browser with JEV (TypeSafe System One): ~100 ms typed decisions, calibrated confidence, human-in-the-loop questions, safety gates and a debug trace.

An open-source AI safety stress-test and intelligence hub for leading models and guardrails, including Jev, ChatGPT, Claude, DeepSeek, Kimi, Gemini, Llama Prompt Guard 2, Grok, GLM, WildGuard, ShieldGemma, Granite Guardian and Shieldstral.

Vietnam - Content safety guardrails for AI chatbots: input, output and conversation checks over one policy file with Jev

Comment moderation console adapted for TypeSafe Jev - spam/abuse/borderline routing at ~$20 per million comments