Agent guardrails with decision models

Pi extension that turns what you say into structured runtime constraints and checks every side-effecting tool call against them.

Pi extension using Jev as a tool-call gate, a bash output judge and a jev_ask tool, in shadow mode by default and failing open.

Auto mode for Pi: deterministic policy first, then Jev judges escalated bash/write/edit calls, failing closed when undecidable.

Lean Pi supervisor: zero-token regex guardrails, done-check verification, per-turn routing and a Jev/Clef semantic slop gate on diffs.

Pi extension that checks proposed code edits against repository Markdown rules with Jev, reporting rule text and line ranges.

Rule-engine layer for Pi where Jev answers closed questions (destructive? leaked key? done verified?) with packs and fail-open actions.

Pi extension that auto-approves routine Bash commands via a System One model and escalates flagged ones, including obfuscated, to a human.

Minimal Pi permission gate: deterministic rules first, then a classifier answers allow, ask or deny per gray-zone tool call; fails closed.

Pi guardrails on pi-typesafe that steer the agent: Jev checks rule breaks, irreversible actions, stuck loops and unverified done claims.

Pre-execution hook for coding agents that lets shell commands run, asks a human, or blocks them, with optional Jev classification.

Rust CLI and Claude Code PreToolUse hook that scores one tool call as Auto, Review or Escalate using a typed decision backend.

Opinionated Pi coding-agent setup with scope control and an improvement loop, plus an optional experimental Jev advisor layer.

Tool-call firewall for Claude Code hooks and MCP servers: Jev answers seven risk questions and code maps them to allow, ask or deny.

Hackathon agent monitor that scores chains of sandboxed agent actions with Jev and escalates through a fixed L1-L5 containment ladder.

Pi coding harness where any LLM writes and Jev returns typed reflex scores that policy code turns into allow, confirm, block, warn or steer.

Discontinued DeepSeek Harness decision layer whose policies let Jev only narrow permissions and fail closed; successor is jevcore.

Pairs Jev with the get-fable lifecycle harness to give coding agents fast tool and skill routing and verbatim-preserving compaction.

Vercel Labs' tiny Zig coding agent, listed for a typesafe_permission_reviewer builtin that runs permission decisions through Jev.

Grok Bot preflight layer: local hard gates first, then Jev judges ambiguity and returns proceed, scope_down, ask_human or block.

Hermes Agent approvals provider where one Jev request answers six typed questions and code derives APPROVE, DENY or ESCALATE.

Python agent development kit for building System One pipelines with Jev for guardrails, routing and PR triage.

Example of Jev as a before-tool-call risk gate in a Google ADK agent, blocking tools when P(risky) passes a threshold.

Python guardrail with a Streamlit dashboard that sends agent tool calls to Jev and allows, blocks or asks for clarification.

Calibrated alignment verifier that checks LLM responses and agent plans for sycophancy, deception, overreach and dark patterns with Jev.

Offline-by-default demo of six Jev decision workflows: model tiering, prompt screening, tool-call risk, reranking and more.

OpenAI-compatible streaming proxy that holds tokens in a buffer and releases them only after a Jev check passes, cutting leaks pre-screen.

Stdlib Python primitive for asking N items K typed questions in bounded, fail-open Jev requests, refusing out-of-set answers.

Strict-allowlist shell harness where Jev walks a TOML decision tree to pick a bash command, flags and paths, confirming writes.

On-device Android agent that taps any app, with an allow/confirm/block safety gate optionally backed by Jev or Laya.

System 1 ingestion gate that asks Jev intent, risk and sensitivity in one call, then blocks, answers deterministically or passes to an LLM.