AgentDock

AgentDock is a Mac application that displays running AI coding agents as screen-edge status dots and collects their prompts in a single inbox.

AgentDock is a Mac application designed for developers who run multiple AI coding agents like Claude Code and Codex simultaneously. It places a discreet strip of status dots on the edge of the screen, indicating whether agents are working, awaiting input, or finished. Questions and permission prompts from all supported agents collect in a centralized inbox for quick answers with an undo window. Users can trigger messages from any application using a double option key shortcut or view turn timers and token usage in hover cards. The application runs locally on macOS, requiring no user account or sign-in, and processes recaps on-device using Apple Intelligence. It supports terminals and multiplexers such as cmux, tmux, iTerm, WezTerm, kitty, and herdr.

AgentDock at a glance

Best for: Software engineer
Runs on: macOS
Pricing: 7-day free trial; $19 one-time purchase for up to 3 Macs
Released: October 2026
Category: AI coding tool
Website: agentdockapp.com

Key features

Displays agents as screen-edge status dots
Collects questions and permissions in one inbox
Sends messages from any app using double option key
Shows model, turn timer, and token usage in hover cards
Provides on-device recaps of finished work
Extends interface to iPhone browsers on the same network

Use it for

Answering permission prompts from multiple coding agents in one inbox
Monitoring working, waiting, and finished agent statuses from screen edges
Sending messages to any coding agent from outside its terminal window
Reviewing daily token usage and model stats via hover cards

Alternatives to AgentDock

Agent OS — Manage AI coding sessions side by side and check on them from your phone.
barn — Monitors multiple AI coding agents on your Mac and jumps to the exact tab when needed.
Coucou — It provides open-source, real-time agent monitoring and approvals directly from the screen notch.
agent-device — Token-efficient snapshots and semantic references for native mobile and TV apps.
Agent Manager — Run multiple AI coding agents side by side with live status, quick prompts, and diff review.
CodePilot — Consolidates multiple AI agents and local model connections in one workspace.
Agent Relay — It carries turns between different AI agents and stops for human decisions.
ManyAgents — Runs multiple AI coding agents locally on your Mac with dedicated worktrees and live previews.
Composio — Equips AI agents with tools, managed user auth, and sandboxes across 1,500+ apps
AGENTS.md — It separates agent-specific instructions from human-focused README files.
Agent Notifications — It removes manual polling and injects CI results straight into the agent loop.
abtop — Monitor multiple AI agent sessions and detect orphan ports from a single screen.

All AgentDock alternatives