Eclipse Enclave provides secure, isolated execution environments for autonomous AI agents working on software projects. It lets developers, teams, and organizations run AI coding agents, and other agentic tools with broad, productive autonomy, while constraining their effect on the host system, the network, and other concurrent work. Eclipse Enclave is not an AI agent and not an IDE. It is the runtime, the isolation layer, and the management surface around agents. It treats agents as untrusted, replaceable workloads that should be observable, sandboxed, and auitable by default.