FAQ
What is AgentRidge, in one line?
A desktop app that runs authorized penetration tests autonomously — a brain agent that plans and delegates, with specialized sub-agents doing recon, exploitation, and analysis. See How it works.
Which platforms are supported?
macOS 12+ (Apple silicon and Intel), Windows 10 (build 17763+)/11 64-bit, and 64-bit Debian-based Linux (Ubuntu 22.04+). See System requirements.
Do I need a powerful machine?
A reasonably capable one. 8 GB RAM and 4 cores is the floor; 16 GB RAM is comfortable. You also need about 25 GB of free disk for the full runtime. The reasoning engine runs locally, which is why. See Local-first design.
Why is the first launch a large download?
To keep the installer small, the local runtime is downloaded once on first launch and reused afterward. It isn't re-downloaded for each mission or app update.
Does my engagement data leave my machine?
The heavy lifting — reasoning, tools, findings, and reports — runs locally. Only account sign-in and licensing talk to the cloud. Your targets, findings, evidence, and reports stay on your workstation. See Local-first design.
What's the difference between Auditor and Attacker mode?
Auditor is proportionate and compliance-oriented (minimal proof of concept). Attacker is maximum-depth adversary simulation (full exploitation and chaining). See Engagement modes. Attacker mode is Pro-only.
Is it really autonomous, or does it just suggest commands?
It's autonomous. The agent runs the full path end to end and pushes each lead until the engagement's definition of done is met. You can watch every command in the Activity stream.
How does it avoid the usual AI refusals for offensive work?
The agents are unlocked to operate as an authorized red-team operator, so they engage with real exploitation tasks — while still requiring a written mandate and staying in scope. See Autonomy & agent unlocking.
Can it damage the systems it tests?
It's built to avoid irreversible harm — no data wiping, ransomware-style behavior, or denial-of-service that risks production (unless scope explicitly allows). In Auditor mode it stays especially light.
What does a mission produce?
A live attack-surface map, a vulnerability register of confirmed findings with evidence, and a client-ready PDF report.
How much does it cost?
There's a free trial (Auditor mode, limited monthly audits and chats). Pro is €2,000 / year and unlocks Attacker mode, LAN discovery, detailed audits, the full runtime, and unlimited chat. See Plans & licensing.
Can I use it on any target?
Only on systems you are authorized to test. Every mission requires a confirmed written mandate before it launches. See Authorized use.
Who makes AgentRidge?
Azuro Software. AgentRidge is a white-label product; inside the app the components are named Agent Ridge Principal, Ridge Sub-Agents, and the Ridge inference engine.
Next: Authorized use →