repoGitHubTrust 82 · PrimaryPublished yesterdayLive · 8h ago
Th0rgal/sandboxed.sh
Safe runtime for autonomous on-chain AI agents: isolated sandboxes, Library skills, encrypted secrets, and OKX read-only security checks.
Lineage graph
Paper → model → repo connections mined from source citations (Tier-1 exact match).
Covers
Implements
Covers (incoming)
Implements (incoming)
Related across the graph
newsthe trust layer is the real productnewsWhy AI Agents Need IsolationnewsCrypto exchange OKX wants AI agents to hire and pay each othernewsI couldn't tell what an AI agent was allowed to do without reading its code, so I built a Dockerfile-shaped way to declare itnewsSecuring the future of AI agentspaperDistributed Attacks in Persistent-State AI ControlpaperAgent-Native Immune System: Architecture, Taxonomy, and EngineeringnewsHow to Secure AI Agents With Container Sandboxing - HackerNoon
