An AI security tool that tests code instead of refusing
Dimitrios from Cosine introduced ArgusRed CLI. He says the free read-only scan can be run now, while pen-test mode needs written authorisation because it tests real systems. The post says the tool uses its own model that was post-trained for offensive security, not just a wrapper around a general model. It also says a multi-agent swarm splits the work and combines the results into one report.
Key points
- The read-only scan is free and available to try now, according to the post.
- pen-test mode is gated behind written authorisation because it runs live offensive testing.
- The maker says this is its own model, not a wrapper around a general AI model.
- The model was post-trained for offensive security tasks.
- A multi-agent swarm divides the work and creates one final report.
Quick term guide
- read-only scan
- A check that looks for issues without changing code or systems.
- pen-test mode
- A mode that tests a system like an attacker would, with permission.
- authorisation
- Official permission to do something.
- multi-agent
- A setup where several AI agents each handle a different subtask and work together to complete a larger goal.
- agent swarm
- Multiple AI agents working at the same time, each assigned a different task.
- Solo makers
- People who build and launch their own products or services entirely on their own.
- AI coding tools
- Programs like Claude, Cursor, or ChatGPT that write code for you when you describe what you want in plain language.
- AI coding tool
- Software that uses AI to help write, edit, or explain code.