Awakado

Product Introduction

1. Definition: Awakado is a macOS menu-bar utility application designed to intelligently manage system sleep states during AI-assisted coding sessions. It functions as an agent-aware activity monitor and sleep inhibitor.
2. Core Value Proposition: It exists to prevent macOS from entering sleep mode during active AI agent execution, thereby eliminating task interruption, while automatically allowing the system to sleep when work is complete to conserve energy. Its primary value is in providing hands-free, context-aware sleep management for AI coding workflows.

Main Features

1. Agent-Aware Sleep Inhibition: Awakado dynamically monitors the activity status of 21 supported AI coding agents (e.g., Claude Code, Cursor, Gemini CLI). It uses a hybrid detection method: for eight agents (Claude Code, Codex, etc.), it installs a lightweight local hook for precise state reporting; for thirteen others (OpenCode, Aider, etc.), it analyzes live process activity. The system only asserts a "caffeinate" claim while an agent's status is "working," and releases it when agents are "waiting," "finished," or "stalled."
2. Lid-Closed Operation (Pro Feature): This feature bypasses the default macOS behavior where closing the lid triggers immediate sleep. It utilizes a privileged system helper (requiring a direct download, not the App Store version) to maintain system wakefulness with the display powered off. Built-in safety protocols (non-disablable battery floor and CPU temperature limits) automatically override this to initiate normal sleep if thresholds are met, ensuring hardware safety during unattended operation.
3. Usage Limit Reset Wake (Pro Feature): Specifically for Claude Code, Awakado implements an automated sleep/wake scheduler. Upon detecting a "usage limit reached" state, it allows the Mac to sleep, programs a system wake event for just before the limit resets (e.g., 4:58 AM for a 5:00 AM reset), and resumes keeping the Mac awake for the agent to continue autonomously.
4. Manual & Hybrid Control Modes: Provides manual overrides that operate concurrently with Agent mode. Users can set a keep-awake duration (30 min to 8 hours), a specific end time, or an indefinite period. A configurable cool-down period (2, 5, 15 min, or off) maintains wakefulness after agents finish to accommodate follow-up prompts.
5. Granular Agent Management: In the Agents tab, users can toggle "Keep the Mac awake" on a per-agent basis. An agent set to "tracked-only" will appear in the activity log and display its status but will not contribute to preventing system sleep, allowing for monitoring without energy impact.

Problems Solved

1. Pain Point: Mid-task sleep interruption during long AI coding sessions. Standard macOS energy settings or generic keep-awake apps cannot distinguish between an AI agent actively executing commands (e.g., running npm test) and being idle awaiting user input, leading to either premature sleep or unnecessary prolonged wakefulness.
2. Target Audience: Developers and engineers who integrate AI coding agents into their daily workflow. This includes full-stack developers using Claude Code for refactoring, data scientists employing AI for script generation, and DevOps engineers running automated tasks via Cursor or Aider. The Pro version targets professionals who run overnight jobs or need to close their MacBook lid while an agent processes a large task.
3. Use Cases:
   • Running a full test suite or complex build process initiated by an AI agent.
   • Allowing an agent to iteratively debug and fix code across multiple files unattended.
   • Processing large codebase migrations or refactoring tasks overnight.
   • Continuing a coding session seamlessly after a Claude Code usage limit resets without manual intervention.
   • Keeping a Mac mini or Mac Studio awake for scheduled agent tasks without monitor interaction.

Unique Advantages

1. Strengths & Limitations (Pros & Cons):

   • Pros:
     • Precision: Sleep management is directly tied to actual computational work, not just application foreground state.
     • Safety-First Design: Non-optional battery and temperature limits mitigate risks of lid-closed operation.
     • Comprehensive Integration: Support for 21 agents via multiple detection methods covers a broad ecosystem.
     • Clear Licensing: One-time Pro fee for 3 Macs with lifetime updates is transparent compared to subscription models.
     • Privacy-Focused: All agent state monitoring occurs locally; no code or prompts are transmitted externally.
   • Cons:
     • Platform Lock-in: Exclusively for macOS 14+, no Windows or Linux support.
     • Helper Dependency: The core Pro feature (lid-closed mode) requires a system helper incompatible with the sandboxed App Store environment, necessitating a direct download and manual security approval.
     • Agent Dependency: Effectiveness is contingent on Awakado's detection capabilities; a novel or heavily customized agent might not be recognized.

2. Key Alternatives & Differentiation:

   • Amphetamine: A popular general-purpose keep-awake utility. Differentiation: Amphetamine triggers are based on app presence, timer duration, or system conditions. Awakado's trigger is the runtime state of specific AI processes, offering finer, automated control that Amphetamine cannot replicate without complex manual scripting.
   • Built-in Agent Features (e.g., Claude's "Keep computer awake"): Often limited and agent-specific. Differentiation: Claude's setting only works with the lid open and may have gaps. Awakado provides a unified, reliable, and cross-agent solution with additional safety and scheduling features (lid-closed, limit reset).
   • Terminal Commands (caffeinate): Manual use of caffeinate -d prevents sleep for display activity. Differentiation: This is a blunt instrument that requires manual execution and termination. Awakado automates the entire lifecycle, integrates with multiple agents, and adds a GUI with logging, safety controls, and status visibility.

Frequently Asked Questions (FAQ)

1. Is Awakado safe to use with my MacBook lid closed? Yes, when using Awakado Pro. It is engineered with mandatory safeguards: it enforces a user-set minimum battery level (e.g., 20%) and respects macOS's internal temperature sensors. If either limit is reached, Awakado immediately stops preventing sleep, allowing the system to cool down or preserve battery. The display is also deactivated when the lid is closed.

2. How does Awakado detect if an AI agent is working versus waiting? It uses a dual-method approach. For integrated agents like Claude Code, it uses a direct hook that reports explicit states ("working," "waiting"). For other agents, it performs live process analysis, monitoring for CPU activity, subprocess spawns (like git, npm, python), and I/O associated with the agent to infer an active working state.

3. Can I use Awakado Pro on both my personal and work Mac? Yes, a single Awakado Pro license is valid for up to three (3) active macOS devices simultaneously. You can manage your licensed devices within the app's License section, where you can deactivate one Mac to free up a slot for another, facilitating use across personal and work machines.

4. What happens if my AI agent gets stuck in a loop? Will Awakado keep my Mac awake forever? No. Awakado includes a "stalled" state detection. If an agent is consuming resources but not making progress (based on its internal heuristics), it will be marked as stalled. After a period, Awakado will cease keeping the system awake for that agent, allowing the Mac to sleep according to your normal energy settings.

5. Why should I use Awakado instead of just changing my Mac's System Settings to never sleep? Manually disabling sleep in System Settings leads to wasted energy, unnecessary wear on components, and security risks if you walk away. Awakado provides an automated, context-sensitive solution that ensures your Mac sleeps the moment your AI-powered work is actually complete, optimizing for both productivity and energy efficiency.