You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Run AI agents in a sandbox that restricts what they can execute, read, write, and reach on the network. Box combines OS isolation with default-deny Dogwood policies and credential injection that keeps secrets outside the agent. Written in Rust. Supports macOS on Apple silicon and Linux.
BenchBloom - a forensic herbarium for benchmark contamination: exact n-gram/Jaccard and containment scans in pure Python, plus a C++17 MinHash engine for finding benchmark items leaked into training corpora.
Sealed-lab research harness for emergent multi-agent (LLM) coordination and containment — replicates the dynamics of the 2026 OpenAI–Hugging Face incident against self-hosted, intentionally-vulnerable targets. Air-gapped by design. Defensive / AI-safety only.
An immersive web-based simulation of the SCP Foundation database with realistic terminal interface, security clearance system, and atmospheric design. Access SCP files, personnel records, and interact with a command-line console in this fan project.
A runtime-enforced containment layer for AI agents: detect, prove, halt, contain, map, and recover without trusting the agent to enforce its own boundaries.