Secure cloud sandboxes for running AI-generated code and autonomous agents.
Open SourceAgent SandboxesCode Execution

Herdr is an open-source runtime for supervising coding-agent sessions across machines. A background service keeps terminal work alive through disconnects and presents running, blocked and completed agents as an attention queue, with terminal, SSH and programmatic clients. Teams should still isolate execution, restrict credentials and review community plugins; repository stars and downloads are popularity signals rather than reliability guarantees.
Secure cloud sandboxes for running AI-generated code and autonomous agents.
Code-first infrastructure for durable AI agents and long-running workflows.
A universal long-term memory layer for personalized AI agents.