Agent‑run Provides a Sandboxed Environment for Coding Agents

Agent‑run is an open‑source tool that runs coding agents in sandboxed containers. It targets developers who need to test autonomous code generation safely. The repository includes scripts to launch and monitor the isolated environment. Sandboxing prevents agents from affecting the host system or network. The project supports customizable resource limits and execution timeouts. Users can integrate the tool into CI pipelines for automated testing. Documentation outlines setup steps for common operating systems. The code is available on GitHub for community review and contribution.