Proactive Claw

What This Skill Does

Proactive Claw serves as the flagship proactive execution engine for the OpenClaw ecosystem. Designed to act as a personal assistant, it continuously monitors your calendar and workflow patterns to optimize your time. Unlike standard automation tools that execute blindly, Proactive Claw operates on a collaborative principle: it analyzes upcoming events, identifies scheduling conflicts, suggests necessary preparation time, and proposes buffers for deep work. It leverages local-first architecture to ensure that your data remains private, processing all scheduling logic through a scoring model that learns your specific preferences and feedback over time. By default, every suggestion is presented to you for approval, ensuring you retain full control over your calendar, follow-ups, and daily agenda.

Installation

To integrate Proactive Claw into your environment, follow these steps:

1. Execute the readiness check: bash scripts/setup.sh --doctor.
2. Obtain the necessary integration dependencies for your platform using bash scripts/setup.sh --print-install-cmd [service] (e.g., google or nextcloud).
3. Establish the calendar handshake: bash scripts/setup.sh.
4. Perform a test run with bash scripts/quickstart.sh.
5. Deploy the engine using the installation command: clawhub install openclaw/skills/skills/googlarz/proactive-claw.
6. Start the background service using python3 scripts/daemon.py.

Use Cases

• Deep Work Protection: Automatically detecting high-intensity tasks and inserting non-negotiable focus buffers before meetings begin.
• Preparation Management: Suggesting 15-minute 'prep blocks' before client meetings based on the nature of the calendar entry.
• Follow-up Automation: Reminding you to schedule follow-up emails immediately after a meeting ends, reducing administrative slip.
• Personalized Scheduling: Learning that you prefer mid-afternoon meetings over early morning ones and prioritizing accordingly.

Example Prompts

• "Proactive Claw, please analyze my schedule for the next three days and ensure I have at least two hours of deep work time per day."
• "I have a busy day tomorrow. Suggest a strategy to move my non-critical syncs to Friday to clear up my Thursday schedule."
• "Add a 15-minute buffer before my 2 PM marketing review to allow for document preparation."

Tips & Limitations

To maximize effectiveness, provide honest feedback on every suggestion; the model improves significantly when you accept or reject its proposals. Note that while everything can run locally, the quality of suggestions is highly dependent on the local scoring model selected. Always review suggested calendar modifications during the first week of deployment to ensure the engine has calibrated to your specific style and professional constraints.