Safety Python Packages

nemoguardrails

NeMo Guardrails is an open-source toolkit for easily adding programmable guardrails to LLM-based conversational systems.

cpex

A policy enforcement runtime for AI agents

skjold

Security audit Python project dependencies against security advisory databases.

safetext

Fast profanity word, curse word, swear word, bad word filtering tool for English, Spanish, Chinese, Turkish and more.

rigel-hpl

High-Level Property Specification Language

acgs-lite

Constitutional governance infrastructure for AI agents — the missing safety layer between your LLM and production

socratic-morality

Constitutional AI Governance Framework - Grounded in Socratic Philosophy

agentarmor

The full-stack safety layer for AI agents. Budget limits, prompt injection shields, PII filtering, output firewalls, and hooks — in 2 lines of code.

phionyx-core

Phionyx Core SDK - AI↔AI Mediator and Echo Engine

robometrics

RoboMetrics is a pip-installable Python library for computing robotics trajectory, prediction, temporal drift, safety, comfort, coverage, calibration, physics, and diversity metrics from NumPy arrays and simple CSV/JSON trajectory files.

loom-agent

Python agent framework with AgentHarness: planner → generator ⇄ evaluator loops for long-running tasks—plus sessions, tool/safety policies, and Claude, GPT, or Gemini.

agent-integrity-proto

Agent Integrity Protocol — real-time thinking block analysis for AI agent alignment

fmdtools

System Resilience Modelling, Simulation, and Assessment in Python

humanproof

Human-in-the-loop checkpoints and approval gates for AI agents

hpl-specs

High-Level Property Specification Language

krisis

A clinical evaluation framework for large language models.

normsync

Norm monitoring and violation detection for AI agent actions

cato-cli-py

Portable sandbox for secure command execution

jmo-security

JMo Security Audit Suite (terminal-first, multi-tool, unified outputs, multi-target scanning)

gpt-scrolls

A collaborative collection of open-source safe GPT-3 prompts that work well

agentos-platform

"The Operating System for AI Agents. Build, Test, Deploy, Monitor, Govern."

dsrl

Datasets for Offline Safe Reinforcement Learning

dbveil

A local read-only, PII-redacting proxy that lets AI agents query your database safely.

model-action-protocol

MAP (Model Action Protocol) — Cryptographic provenance, self-healing critic, and state rollback for autonomous AI agents. 2.5k lines of TypeScript, 60+ tests, MIT.