Latest

On March 1, 2026, I competed at the USAPL Winter War powerlifting meet held at Impact Zone, Norwood. It was an incredible experience competing alongside strong lifters and testing my strength across squat, bench press, and deadlift.

On February 23rd, [Anthropic published](https://www.anthropic.com/news/detecting-and-preventing-distillation-attacks) something the industry had suspected but hadn't seen documented at this scale. Three Chinese AI labs (DeepSeek, Moonshot AI, and MiniMax) ran coordinated campaigns against the Claude API. They generated over 16 million exchanges through approximately 24,000 fraudulent accounts. The goal was not to steal user data but to steal the model itself.

Aurascape's research team this week published SilentBridge, a class of indirect prompt injection attacks against Meta's Manus AI agent. The attack exfiltrated email, extracted secrets, achieved root-level code execution, and exposed cross-tenant media files via CDN — all three variants scored CVSS 9.8 (Critical): network-exploitable, no privileges required, no user interaction. The user had authorized Gmail and the agent used it exactly as permitted. Vulnerabilities discovered September 2025, Manus mitigated November 2025, coordinated disclosure February 2026.

The Clinejection attack chained prompt injection, cache poisoning, and a credential isolation failure into a supply chain compromise. Here is the attack anatomy and what structural controls actually stop it.

When running OpenClaw locally, maintaining a strong security posture depends on disciplined execution. Many teams begin with permissive settings and attempt to lock things down later, unintentionally accepting risk and facing tougher enterprise reviews. Instead of simply enabling some settings, establishing system invariants that guard against threats like **Indirect Prompt Injection** and **Confused Deputy** attacks is the key.