← Discover
Discover · GitHub · ADR↑ +148 today

Security for enterprise AI agents

Developed by Uber, ADR protects corporate AI agents with observability, security benchmarking and threat detection methods. The software offers a security framework developed to increase the security of artificial intelligence systems.

Updates

  • August 6, 2026: Stars 782 → 1,123, latest release sensor-v1.0.0 (July 31, 2026).

What you get

  • It monitors the actions taken by AI agents, allowing you to understand what they do and why they do it.
  • Tests agent security with over 300 missions against 17 different attack techniques.
  • It uses a two-stage architecture to detect suspicious sessions.

Installation

Preparing the detection system
git clone https://github.com/uber/ADR
cd ADR/Detection
uv sync
export ANTHROPIC_API_KEY="..." OPENAI_API_KEY="..."

If you don't write code

🤖 Paste this into your AI agent (Claude Code · Codex · Antigravity)

How can I protect my enterprise AI agents against vulnerabilities? How can I monitor agent activities, detect threats, and evaluate the defense power of my system by making security comparisons using the ADR system? Please explain the steps I need to follow in this process and how to configure the observability features that ADR offers.

Related dictionary terms

Who it is forIt is suitable for security experts and software developers who run enterprise-level artificial intelligence agents and want to ensure the security of these systems.
LicenseApache-2.0

Links

TreScout did not build this tool · we found it in GitHub trends and wrote it up. This page describes the repository as of 2026-08-05: The star count and our text belong to that day, the repository may have changed since. Check the repository link for the current state. This page was machine-translated from the Turkish original · the Turkish version prevails.