Name: AgentPeek
Rating: 4.0 (1 reviews)

AgentPeek lives in your Mac notch and menu bar and shows your Claude Code, Codex, Cursor, Grok, Hermes, OpenCode, and Antigravity sessions in real time, so you can glance up instead of digging through windows. See active sessions, permissions, token usage, and local dev servers at a glance, across terminal and desktop. Your session data stays local on your machine. Native, fast, and built for anyone running AI coding agents all day.

What I like most about AgentPeek is that it solves a very specific problem without trying to become another AI workspace.

When you’re running multiple coding agents across terminals and apps, simply knowing what is active, waiting for permission, or consuming tokens can become surprisingly distracting.

Putting that visibility into the Mac notch/menu bar feels like a natural place for it.

I also appreciate the local-first approach. As agents become more deeply connected to our actual workflows, where session and context data lives is becoming an increasingly important part of the product experience.

Still early in my use, but I like the direction: focused, native, and built around an increasingly real problem for people working with multiple AI coding agents.

As the number of active agents grows, I’d love to see even more at-a-glance prioritization.

For example, clearly surfacing which sessions need attention, which are actively working, and which have been idle for a while could make the overview even more useful for heavy multi-agent workflows.

I also looked at broader multi-agent workflows and tools that manage coding agents from dedicated dashboards or terminals.

What appealed to me about AgentPeek was the lighter-weight approach: rather than becoming another workspace, it adds visibility to the workflow I’m already using on the Mac.

What I like most about AgentPeek is that it solves a very specific problem without trying to become another AI workspace.

When you’re running multiple coding agents across terminals and apps, simply knowing what is active, waiting for permission, or consuming tokens can become surprisingly distracting.

Putting that visibility into the Mac notch/menu bar feels like a natural place for it.

Still early in my use, but I like the direction: focused, native, and built around an increasingly real problem for people working with multiple AI coding agents.

As the number of active agents grows, I’d love to see even more at-a-glance prioritization.

I also looked at broader multi-agent workflows and tools that manage coding agents from dedicated dashboards or terminals.

What appealed to me about AgentPeek was the lighter-weight approach: rather than becoming another workspace, it adds visibility to the workflow I’m already using on the Mac.

AgentPeek

Your coding agents, in the Mac notch

Your coding agents, in the Mac notch

AgentPeek Launches

AgentPeek Launches

What's great

What needs improvement

vs Alternatives

What's great

What needs improvement

vs Alternatives