Systems
Open SourceAgent dashboard (open source)

Agent Commander

Multi-host dashboard for AI agent sessions. Real-time terminal streaming, attention queue, scoped memory, approval flows. The open-source version of how I run agents.

Most agent tools are built for one agent on one machine. This one assumes you're running a fleet across several boxes and need to triage them.
System Card
categoryAgent infrastructure
statusOpen Source
page/agent-commander
1

What it does

Streams terminal output from Claude Code (and similar) sessions across machines. Surfaces approval requests in an attention queue so I can decide what to authorize in 30 seconds. Records scoped memory so an agent picks up where it left off after a context reset.