Home / Blog
Field notes from the engagements
Blog
What we find, how it is exploited, and what actually fixes it. Written by the people who ran the test.
AI and LLM Security
Exposed AI Agents Are the New Shadow IT – Except They Can Act
175,000 Ollama servers exposed to the internet, 8,000 MCP servers wide open, and 31% of those examined answering with no authentication at all. Same film, new cast -…
Read itCloud Security
Attackers Don’t Break Into Your Cloud. They Log In.
No exploit, no zero-day, nothing that trips an alert. A Cognito refresh token can be valid for ten years, and CloudTrail sees everything while understanding nothing.
Read itAI and LLM Security
Prompt Injection Is the SQL Injection of the AI Era
In 2005 everyone built websites and few asked what happens if someone injects code into a text field. Today everyone builds AI agents, and few ask what happens…
Read itCloud Security
There Is No Safe Starter Kit. There Is Only One Nobody Has Tested Yet.
A starter kit that generates IAM roles with wildcard permissions on every resource. AWS's fix was to update the documentation.
Read itSupply Chain
500 AI-Written Attacks, 90% Stopped by a Default Setting
An AI tool analysed 500 repositories and wrote a custom payload for each. Then it waited for GitHub Actions to run them automatically - which GitHub does not…
Read itAI and LLM Security
AI Wrote the Code. The CVE Is Still Yours.
Researchers traced real CVEs back through git blame to the commit that introduced them, then checked whether AI wrote it. Code that arrives looking perfect is exactly the…
Read itEngagements and Process
I Run a Company of Pentesters, and I Watch AI Bots Doing Their Job
Faster, cheaper, and far more of it. The question is not whether AI replaces pentesters - it is which pentesters survive, and what they will need to be…
Read itAI and LLM Security
Is Your AI Secure? The Dark Reality of LLM Vulnerabilities
Cyber attacks targeting large language model (LLM) based applications are increasing in both frequency and sophistication. Misconfigured chatbots, exposed APIs, and weak integrations with web, cloud, and mobile…
Read itSupply Chain
A Regex Without Anchors, and a Race to Register a GitHub Account
A webhook filter checked ACTOR_ID with a regex that had no ^ or $. Substring match was enough. This is not a regex mistake - it is a…
Read itApplication Security
What Is Real-Time Penetration Testing And Why It Matters
One quiet vulnerability can be tomorrow’s headline, and the clock is always in the attackers’ favor. IBM’s 2025 Cost of a Data Breach Report just discovered the average global…
Read itTell us what the system does and what worries you.
If a penetration test is not what you need yet, we will say so.