A living map of the AI agent security ecosystem.
-
Updated
Jun 12, 2026
A living map of the AI agent security ecosystem.
A Serverless Security Library for Developers. Regain Control Over Your AWS Lambda & Google Cloud Functions Runtimes.
Circuit breaker for production PostgreSQL. Native extension that can ALLOW, WARN, or BLOCK dangerous SQL based on live database state. Apache 2.0.
🛡️ Proactive Runtime Injection Shield & Monitor for OpenClaw — injection scanning, tool execution guard, DLP, and file integrity monitoring
Runtime integrity guard: detect and block Linux page cache tampering (Copy Fail, Dirty Pipe, Dirty Frag) via O_DIRECT + fanotify. Includes 7 host-side exploitation PoCs.
A C++ runtime protection framework for Windows, focused on anti-tamper mechanisms, executable protection, and security hardening using MSVC and CMake.
🛡️ The CrowdStrike for AI Agents — Runtime protection, threat detection & security monitoring for LLM agents. Supports LangChain, CrewAI, AutoGen, OpenAI. Python • JS • Rust • Go • Ruby • .NET
Security Research Project: A B.Tech minor project built upon an Android calculator application, extended to implement and evaluate security mechanisms such as anti-debugging, root and emulator detection, runtime protection, and code obfuscation. The work focuses on analyzing Android app resilience against tampering and reverse engineering.
Dhi - Runtime Intelligence & Protection System. Sanskrit: Intellect | Perception. Kernel-space eBPF security monitoring for AI agents.
Control plane for enterprise AI security: inline runtime guardrails, automated red teaming with auto-remediation, AI asset posture (AI-SPM), and compliance governance. Python control plane + Go inline agent.
Runtime protection for AI coding agents. Intercepts tool calls, scores them against the task, blocks prompt injection attacks before execution.
Top Cloud Workload Protection Platform CWPP (Opensource) Star if you like it!
Simple low-level Android antivirus
Top AI Runtime Protection (Opensource) 🌟 Star if you like it! 🌟
Raspire Inc. (RASPIRE) is a mobile application security company offering AI-powered, no-code runtime application self-protection for Android and iOS apps. Its SHIELD / XTD platform applies configuration-driven defenses that protect every release from code tampering, reverse engineering, hooking/injection, MITM interception, rooted or jailbroken…
Runtime guardrail for AI-generated Python code. Blocks rm -rf, curl | bash, and unapproved network calls before execution.
SentinelGuard-SDK is a future-ready, hardware-agnostic security library designed to implement smart control and safety protocols for electronic devices and advanced machinery. Built with a 'Security-by-Design' approach, it utilizes asymmetric cryptography, geofencing, and remote telemetry to ensure authorized-only access.
Enforces strict syscall whitelisting and egress IP cidr blocks for sandboxed autonomous code execution environments | GenPark: https://www.genpark.ai | LinkedIn: https://www.llinkedin.com/company/92955550
Security scanner and runtime protection for MCP (Model Context Protocol) servers
A clone of an opensource project where I was a major contributor for Reblaze in 2020. See also the talk https://youtu.be/6_8CkDWz3XQ that was inspired by this work.
To associate your repository with the runtime-protection topic, visit your repo's landing page and select "manage topics."