An independent research notebook on AI engineering, agents, models and the systems around them.https://julin.ai/atom/tags/infrastructure/index.xmlJu Lin2026-09-29T00:00:00+13:00https://julin.ai/2026/09/29/claude-agent-runtime/2026-09-29T00:00:00+13:002026-09-29T00:00:00+13:00<p>Different Claude tools run agents in different places. The choice shapes what the agent can do.</p> <h2 id="claude-claudeai">Claude (claude.ai)</h2> <p>The agent runs inside a gVisor container on the server side. No local setup required—everything runs remotely. The container is ephemeral; each session starts fresh and leaves no traces.</p> <h2 id="claude-code">Claude Code</h2> <p>The agent runs on your machine. It can execute every command and access every file that your logged-in user can access. For advanced users concerned about blast radius, start a sandbox first (like a microVM) and run Claude Code inside that isolated environment.</p> <h2 id="claude-cowork">Claude Cowork</h2> <p>The agent runs on a VM. On macOS, it uses the Virtualization framework; on Windows, it uses HCS. The VM has its own Linux kernel and filesystem. But the agent loop itself runs outside the VM—only code execution happens inside.</p> <hr> <p>Each design is a tradeoff. Server-side isolation is safe but ephemeral. Local execution is powerful but risky. VM-based execution gives you isolation with some local access.</p>