Havuz

Code review via the change impact measure

The vendor-neutral protocol for AI agent safety & security — and the neutral benchmark that ranks the vendors.

不死鸟 Phoenix — Hermes Agent 插件：路由分档/风险防线/自愈/存档点提醒/审批策略自适应，官方钩子接入，不改 Hermes 核心代码

sofagent is an open-source FDE Harness for AI coding agents — on entry, write your business judgment into files (workflows, ontology, AI nod

🤖 LLM Engineering Roadmap — Complete Developer Guide

AI 执行与监察体系（Agent Oversight Framework）：分层治理规范 + L2 运行时监察参照实现——证据分级、自校正、边界诚实。A layered governance spec for AI agents, plus a runnable referenc

Grade every agent, scraper and model in a claim's chain — provenance, trust scoring and audit evidence for LLM pipelines

Chinese-first LLM jailbreak benchmark & red-team toolkit | 中文LLM安全评测框架 | by POJIA.AI

Protecting digital communications against audio deepfakes with real-time AI. This system analyzes audio streams to detect synthesized voices

A Socratic supervision layer for AI coding agents.

Open-source AI model evaluation and benchmarking framework for LLMs (OpenAI, Ollama, Claude, Gemini)

AI Firewall & LLM security toolkit - protect your AI applications from prompt injection, jailbreaks, PII leakage, and adversarial attacks

End-to-end pipeline for seeing how LLMs actually process your prompts. Capture attention across every layer, render heatmaps and cooking cur

Every way an LLM can break one SDK to stop it. Injection. Hallucination. Leakage. Handled.

Open database of system prompts extracted from frontier LLMs using JustAsk

Biologically-grounded adversarial training platform: cyclic Wake/Dream/Nightmare/Compress phases that accumulate model robustness without ca

🎱 AI Agent Governance Framework — Constrain how AI Agents behave in your project. pip install pattern8

High-fidelity Claude Fable 5 (Mythos) environment emulation and automated multi-agent jailbreak (Pack Hunt) research laboratory.

Verify what AI coding agents can actually read, write, execute and reach using synthetic canaries.

An unattended maintainer for one repository. It reviews, files, prunes, and merges only against proof that a guard bites.

A modular threat detection library for AI and LLM pipelines. It provides language-agnostic guardrails through a configurable pipeline of loa

Local-first verification layer for AI agents: evidence, provenance, policy gates, human approval, and auditable Trust Receipts before a clai

Reproducible OpenEnv-compatible reinforcement-learning environments for social-media integrity moderation.

Detect hallucinations in LLM responses. Verify every claim against source documents using hybrid STS + NLI. Works with LangChain, LlamaIndex

Free local guardrail for AI coding agents: every tool_call judged on-device (regex + Laya), allow/review/block

1,017 system prompts from real AI products, audited span by span against the eight AISPA assurance dimensions

A custom coding harness for TypeSafe AI's Jev: an LLM proposes, Jev answers narrow questions, code decides, every step leaves a receipt.

A terminal-based, multi-domain support-ticket triage agent grounded in a local support corpus.

Overnight experiment: can an AI agent make money, and can you trust what it reports? The agent produces claims; a separate verifier it canno

Tamper-evident evidence and verification layer for AI agents. Verify any agent's session offline with just openssl. No cloud, no lock-in.

Lightweight RAG provenance middleware. Verifies every claim in an LLM response is grounded in a retrieved source - without an LLM call.

Open-source test harness for AI agents. Stress-test production agents with adversarial multi-turn scenarios in CI

Zero-cost epistemic uncertainty quantification & hallucination detection for LLMs (90,000x faster than Semantic Entropy)

AI makes users productive — but also cognitively lazy. Lucid detects cognitive dependency in real-time and adapts AI responses to keep users

The gate between an AI agent and real money. Runtime authorization for financial AI agents: structured intent, deterministic policy, human a

AI agents now operate with authority. Authority without discipline is how complex systems fail. Nuclear’s control loop, ported to AI-assiste

Myrtia — agent-native red teaming for LLMs and AI agents: a deliberating council of attacker models (propose, cross-critique, Borda vote, ch

⚡ Millisecond System-1 judgement for every tool call in DeepSeek Harness — Jev-powered risk classification & evidence-gated auto-approval. F

A structured workflow system for AI coding agents - harness engineering, execution loop, skills, hooks, and a learning loop. Works with Clau

🔥 CogitoAgent — An autonomous AI agent, cloud-driven but locally executed. It continuously thinks, automatically analyzes, and integrates pr

Audit any agent decision across its past, present, and future, on one typed graph.

RLCDAlignBench: 44 alignment-failure detection benchmarks and code for 'Just Ask Jev' (RLCD zero-shot detector of AI alignment failures)

An open database of real-world AI agent incidents.

Checks every AI agent action against your company's real rules before it runs. Git-native, PR-reviewed.

MCTS (Model Context Threat Scanner) is a local-first security scanner for MCP servers -- static and live tool discovery, multiple analyzers,

Python SDK for LLM guardrails with safety classification, PII detection, prompt injection defense, and grounding checks. Protect AI applicat

Noisegate: a differential privacy gateway that lets an untrusted LLM agent query sensitive data over MCP (Model Context Protocol), with a fo

Parent-operated starter kit for designing a private, child-facing Hermes Agent profile

Prompt level, inference time, AI frameworks to improve safety and alignment.

The execution safety layer for AI agents.