AI guardrails & agent security
3,625 agents · most starred first
AI guardrails & agent security, page 2 of 37
AI Bill of Materials — discover every AI agent, model, and API in your infrastructure
An SSH authentication agent for Cygwin/MSYS to PuTTY's Pageant.
Production-grade, AI-Agent-friendly multi-platform OAuth harvesting and session management framework for AI Gateways.
HASP is a local-first broker for managed secrets in agent workflows.
SecureClaw - Security Plugin and Skill for OpenClaw OWASP-Aligned
The local runtime control layer for OpenClaw/Hermes-agent, PII & sensitive credentials protection.
SAFE-MCP is a comprehensive security framework for documenting and mitigating threats in the AI Agent ecosystem.
Security firewall for agents
Claude Code 免杀 SubAgents
Configurable Multi-layered AI Agentic Safety Framework
1Password for coding agents: find, use, and save secrets without exposing them
Lightweight Agent Detection & Response (ADR) layer for AI agents — guards commands, files, and web requests. Part of Gen Agent Trust Hub.
Agent Security Bench (ASB)
Remote approvals, policy checks, and execution evidence for unattended AI agents.
Halofy is the open access and governance layer for AI agents across your organization. Identity, policy, provenance, audit, and signed erasure.
[USENIX Security 2025] PoisonedRAG: Knowledge Corruption Attacks to Retrieval-Augmented Generation of Large Language Models
Agentic AI EDR for developer workstations and autonomous agent swarms. Build Swarm Detection & Response platforms with Clawdstrike.
A native policy enforcement layer for AI coding agents. Built on OPA/Rego.
Cloud native zero trust security for AI agents run environments
Safety at Scale: A Comprehensive Survey of Large Model and Agent Safety
Nova-Proximity is a MCP and Agent Skills security scanner powered with NOVA
The authentication CLI & Proxy for AI agents. Give agents access, not your credentials.
🔐 Firewall Your Data, Control Agents. Prevent agent data exfiltration. Gain visibility into AI's interactions with your data / systems of record / existing software. https://discord.gg/tXjATaKgTV
🛡️ A curated list of resources on agent skills security: attacks, defenses, frameworks, and benchmarks for securing AI agent tool use and skill ecosystems
API gateway for purpose-based authorization for AI agents. Human approval for tasks, AI-native enforcement for agents.
OpenART is an open-source framework designed to evaluate the safety and robustness of autonomous AI agents in dynamic, long-horizon, and stateful environments. It stress-tests agent runtimes against multi-step state poisoning, privilege escalation, and tool-use vulnerabilities across 10,000+ benchmark scenarios.
DodoGuard: Take AI Agents to Production. Test, validate, and operate AI Agents before and after deployment.for Dify, Coze, FastGPT, n8n, Tencent ADP, ByteDance HiAgent, and more.
Hands-on code, labs, and reference material by Omar Santos for building, securing, and operating AI agents in cybersecurity.
ArmorIQ intent enforcement plugin for OpenClaw AI agents
The open standard for runtime agent control — declarative hooks, policy enforcement, and observability across AI agent frameworks.
Harness is an AI Agent development guardrail Meta-Skill that establishes four layers of defense for any project in one command: knowledge management, architecture constraints, feedback loops, and entropy management.
[NeurIPS 2024] Official implementation for "AgentPoison: Red-teaming LLM Agents via Memory or Knowledge Base Backdoor Poisoning"
Find known secrets in your AI coding agent sessions (Claude Code, Codex, Cursor), then guard the next one with local hooks that trace risky commands to the content that dictated them. Open source, no telemetry.
SeClaw provides a practical foundation for measuring, diagnosing, and comparing security failures in autonomous LLM agents.
Baseline rules files to improve the security of AI-generated code (Claude, Cursor, Copilot + more)
Static security scanner for LLM agents — prompt injection, MCP config auditing, taint analysis. 51 rules mapped to OWASP Agentic Top 10 (2026). Works with LangChain, CrewAI, AutoGen.
Govern: one Decionis verdict before a workflow step runs — a deploy, a migration, an infrastructure change, an agent's action — with a signed Decision Dossier of it. GitHub Actions, GitLab CI, Jenkins; shadow or enforce; no local policy engine.
Open-source adversarial testing engine, SDK, and CLI for AI agents. Runs locally or against the Humanbound Platform.
Your AI's guard dog. Doberman sits at runtime, gating every input, output and tool call to stop unsafe or unintended actions before they execute.
🤫 AI-native secrets manager. Agents use secrets without seeing them.
Runtime security for AI agents: discover agents, map their permissions, and enforce what they can do.
Semia, security audit for AI agent skills.
[ICLR'24 Spotlight] A language model (LM)-based emulation framework for identifying the risks of LM agents with tool use
Red Teaming python-framework for testing chatbots and GenAI systems.
🐾 Authority, not credentials, for AI agents: policy-approved actions, delegation that can only narrow, instant revocation, tamper-evident audit.
Self-hosted execution layer for AI agents: connect any agent to real APIs, scope exactly what it can touch, and keep credentials out of the agent's hands.
Access control for AI agents. Set what Claude Code, Codex, Gemini, Cursor and any MCP server are allowed to do, review risky actions before they run, and keep every action on the record.
Open-source AI security verification for model artifacts, live endpoints, MCP servers, and recorded agent traces. Reproducible evidence for release decisions.
TrinityGuard: A Unified Framework for Safeguarding Multi-Agent Systems
The enterprise gateway for autonomous agents. Identity management, per-channel isolation, credential vault, per-session tamper-evident audit log.
Highflame ZeroID: Autonomous Agent Identity Management System (AAIMS)
Your AI's guard dog. Doberman sits at runtime, gating every input, output and tool call to stop unsafe or unintended actions before they execute.
Runtime Security for tool-using AI agents, with local policies, pre-action enforcement, and forensic audit trails.
Runtime defense against memory poisoning in AI agents (OWASP ASI06). Drop-in for LangChain, OpenAI Agents, AutoGen, CrewAI and mem0.
Protect LLM agents against malicious prompts using JEV
Top 10 for Agentic AI (AI Agent Security) serves as the core for OWASP and CSA Red teaming work
Zero-knowledge credentials infrastructure built for AI agents to operate, not just consume.
Toolkits to create a human-in-the-loop approval layer to monitor and guide AI agents workflow in real-time.
Scan your dev machine for AI agents, MCP servers, IDE extensions, and suspicious packages - in seconds.
ClawBands is a security middleware for OpenClaw AI agents.
IAGA Sentinel sits next to your AI agents and answers the one question the agent itself cannot.
EU AI Act compliance proxy for AI systems. Drop-in HTTP proxy that monitors every AI interaction for regulatory risks, logs to a tamper-proof audit trail, and generates legal-grade PDF reports. 143 rules across 8 Annex III domains. Self-hosted, open-core
Stop AI agents from doing things they shouldn't. The action authorization layer — pre-execution, deterministic, policies included. Drop in. Day one.
The secure gateway connecting AI agents to enterprise systems.
Local guardrails for AI coding agents. Check proposed file changes and shell commands, block risky patterns, and return actionable feedback. Works with Claude Code, Codex, Cursor, Gemini CLI, and more
Security layer for AI coding agents. Works with Claude Code, Cursor, Windsurf, Gemini CLI, OpenCode, Pi Agent and more.
Audit AI agent tool-call transcripts: hash-chain sealing, deterministic replay, and scope overreach checks. Dependency-free Python CLI.
Local-first, hardware-bound key management and encrypted secrets for AI agents.
Advanced prompt injection defense system for AI agents. Multi-language detection, severity scoring, and security auditing.
An open-source engineering governance standard defining trust boundaries for conversational AI agents in high-stakes domains. MIT Licensed.
An open-source engineering governance standard defining trust boundaries for conversational AI agents in high-stakes domains. MIT Licensed.
A Geiger counter for AI agents — one read-only command that inventories every agent, MCP server, plugin, and AI extension on a machine
auth for agents–from the creator of `dotenv` and `dotenvx`
Context bomb strings
Introducing XSafeClaw: The Open-Source Agent Safety Platform from Fudan University
GitHub Agentic Workflows Firewall
Build Secure and Compliant AI agents and MCP Servers. YC W23
A Mac vault that runs authenticated actions for AI agents over MCP. The agent gets the operation, never the key: no command reveals a stored credential, and there is no export route.
See what your AI coding agents are doing. AEGIS 0.25.0-beta: open-source Windows desktop with 8 interface languages, process/file/TCP monitoring, project and skill reviews, Claude Code usage and audit reports. Monitoring needs no agent plugin or subscription. Opt-in action policies and MCP gateways cover selected routes.
The Universal Governance, Risk, Compliance (GRC) Operating System with Integrated Security for Agentic AI, Non-Human Identities, and Swarm Governance. AI SAFE² + AI Sovereignty Maturity Model (AISM), NEXUS-A2A Protocol, FORGE-Act (e.g. Marshal Plan for AI) [Dual License: MIT + CC-BY-SA]
Open-source governance and hard gates for AI coding agents across Claude Code, Codex CLI, and Pi.
The fastest Trust Layer for AI Agents
Auto mode for every coding agent, built on Jev: risk-scores every tool call with session context (deny / ask / allow), flags prompt injection in results, checks skills and plugins. Claude Code, Codex, Copilot, Gemini, Cursor, pi, OpenCode, ACP.
Govern consequential AI agent actions in Docker with deterministic policy, human approval, and signed Decision Dossiers.
Static analysis for AI agent configs, tool descriptions, and system prompts — catches vague tool descriptions, missing stop conditions, and schema gaps before they reach runtime. Zero-LLM, deterministic checks, built for CI.
Runtime enforcement boundary for AI agents: a local sidecar that gates every outbound call against Cedar policies you own. Deterministic, call-level, no model on the hot path
RISE: Red-teaming via Iterative Strategy Evolution (NeurIPS 2026)
The security layer for AI coding agents : Skill governance, safe package recommendations, MCP/tools guardrails, secret protection, runtime policy enforcement and full audit visibility with a self-serve dashboard.
AI classifier for indicating AI-written text
A comprehensive security scanner for Model Context Protocol (MCP) servers that detects vulnerabilities and security issues in your MCP server implementations.
Open Source policy enforcement proxy: Make your agents follow rules.
AI 应用合规网关 · 一行命令体检 AI 项目的「数据出境 / 硬编码密钥 / 个人信息暴露」(网安法·PIPL·等保2.0·数据出境·AI标识),并给出境内模型替代建议;可作运行时防护拦截注入与数据外泄 · 中文优先 · 零依赖 · 开源
Claude Code-style declarative permission rules for DeepSeek Harness: ordered allow/deny/ask rules with tool-name, argument (glob/regex), and workspace-path matching on the tools/pre-execute waterfall, session-log audit, and HMR reload.
SCAM - Security Comprehension Awareness Measure | Open-source benchmark that tests AI agents' security awareness during realistic, multi-turn workplace tasks.
Security gateway for AI agents - credential-isolated API proxying and policy-gated remote execution (conclaves). Reduce the blast radius!
Open source prompt injection protection for Agents calling tools (via MCP, CLI or direct function calling). Detect and defend against prompt injection attacks. 22MB, CPU-only, < 10ms latency.
Temporary access to secrets and private networks for one command, then gone. Same wrap for humans and agents. See which access was used.