A centralized task management and scheduling system that ensures AI agents never miss a commitment or recurring responsibility.
The fastest way to install a skill directly from the registry.
npx clawhub@latest install pulse-todo
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
Copy this prompt to OpenClaw to install it automatically.
Help me install pulse-todo using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Get the raw skill files in a ZIP archive.
Pulse TODO provides a single source of truth for an AI agent's pending work, organizing everything from one-off tasks to complex recurring schedules. By utilizing Openclaw Skills to maintain a workspace-root TODO.md, agents can bridge the gap between human requests and autonomous execution. This framework eliminates the reliance on short-term model memory, replacing it with a robust system of record that integrates directly with cron jobs for reliable long-term scheduling.
The skill is designed around the core principle that everything is a task, differentiated only by its timing and dependencies. It allows agents to intelligently choose what to work on next by aligning task selection with strategic goals defined in documents like SOUL.md or MEMORY.md, ensuring that the agent remains productive during idle moments or heartbeat triggers.
To begin using this skill within your workspace, ensure you have the Openclaw Skills environment configured and follow these steps:
# Create the primary task tracking file
touch TODO.md
# Initialize the standard Pulse TODO structure
echo -e "# TODO\n\n## 📋 Tasks\n\n### Do it myself\n\n### Depends on human\n\n### Waiting on external\n\n## 🔄 Scheduled" > TODO.md
Refer to the setup.md file in the skill repository for specific cron permissions and agent profile integration.
The skill maintains a structured Markdown file at the workspace root named TODO.md. The schema is organized as follows:
| Section | Description |
|---|---|
| Do it myself | Tasks actionable by the agent immediately without external blockers. |
| Depends on human | Tasks requiring input, review, or action from a human user. |
| Waiting on external | Tasks waiting on third-party fixes, PR merges, or external systems. |
| Scheduled | Recurring or timed tasks linked to automated triggers. |
Task Attributes:
repeat: <schedule>: Defines the recurrence pattern (e.g., daily 3:00).once: <timestamp>: Defines a one-time execution window.cron: <job-name>: Links the entry to a specific background job for Openclaw Skills synchronization.Loading
A Model Context Protocol skill for managing Google Search Ads and keyword research directly through an AI agent.

An AI-powered ad management tool for creating, targeting, and optimizing Meta Ads campaigns through natural language interfaces.

A discovery tool for finding and installing modular agent capabilities from the Openclaw Skills ecosystem.

A specialized research skill for analyzing the North Atlantic Treaty Organization's history, membership, and collective defense mechanisms.

An AI-powered regional expert providing real-time data on United States travel, economic shifts, and cultural nuances.

A specialized AI skill that provides real-time encyclopedia-style data and travel information for the Maldives.








































