A sophisticated state management and session recovery system that prevents data loss during subagent timeouts by using checkpoints and scheduled cron resumes.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install smart-wake
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install smart-wake using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is smart-wake?
smart-wake is a critical reliability framework for developers building complex Openclaw Skills that require extended execution times. By implementing a standardized checkpoint-and-resume architecture, it ensures that long-running tasks can survive the inherent timeout limits of subagents. This skill acts as a persistence layer, allowing an agent to save its progress, schedule a wake-up call via gateway cron jobs, and continue work in a fresh session without missing a beat.
At its core, smart-wake solves the challenge of non-deterministic or multi-hour workloads that exceed standard execution windows. Instead of failing when a timeout threshold is reached, the system orchestrates a graceful handoff using state persistence and session spawning. This makes it an indispensable tool for building resilient, enterprise-grade AI agents that must handle heavy data processing, repository-wide scans, or periodic background monitoring.
smart-wake Use Cases
- Managing multi-hour repository analysis or codebase refactoring tasks that exceed default execution limits.
- Automating periodic data polling and synchronization tasks that require background execution over several days.
- Running high-complexity Openclaw Skills during off-hours with automated resumption and progress tracking.
- Ensuring zero state loss for multi-step AI workflows that may pause for external dependency readiness.
- Improving the reliability of agents performing large-scale content generation or batch processing.
How smart-wake Works
- Preflight Estimation: The agent calculates the total task duration and enables the smart-wake protocol if the estimate exceeds 70% of the session timeout.
- Chunked Execution: Tasks are broken down into manageable 5-15 minute segments to facilitate regular state updates.
- Checkpointing: After each chunk, the agent writes its current status, progress percentage, and next actions to a mandatory state contract.
- Pre-timeout Auto-save: As the timeout approaches (within 10-20% remaining time), the agent flushes a final checkpoint and sets its status to waiting.
- Gateway Cron Scheduling: The agent registers a cron wake event with the gateway, passing a payload containing the task ID and resume context.
- Session Resumption: At the scheduled time, a new session is spawned via session_spawn, which reads the latest checkpoint and resumes execution from the exact point of interruption.
smart-wake Setup
To integrate smart-wake into your Openclaw Skills, you must ensure your environment has access to gateway cron jobs and session management tools.
# Verify environment support for session management
openclaw tools list | grep "session_spawn"
Enable the protocol by implementing the mandatory state contract in your agent's memory slot and configuring your gateway to recognize the wakeMode parameter for cron-based resumption.
smart-wake Data Schema & Taxonomy
The smart-wake protocol relies on a strict state contract to ensure idempotency. Every task must maintain the following schema:
| Field | Description |
|---|---|
task_id |
Unique identifier (e.g., smartwake_scan_20260101). |
status |
Current lifecycle state: running, waiting, blocked, done, or failed. |
progress_pct |
Integer representing task completion. |
next_actions |
Array of steps to be performed after the next wake cycle. |
artifacts |
List of generated file paths or data pointers to verify upon resume. |
updated_at |
ISO timestamp of the last successful checkpoint. |
retry_count |
Counter to prevent infinite wake loops (default max 5). |
smart-wake Advanced Features
- Idempotency Guard: Automatically verifies artifacts and checksums upon resumption to prevent redundant processing.
- Dynamic Wake Intervals: Adjusts the cron schedule based on system load or task-specific dependencies.
- Multi-Agent Orchestration: Supports session_spawn for delegating complex sub-tasks while maintaining the parent wake lifecycle.
- Automated Cleanup: Triggers a teardown of all associated gateway cron jobs once the task status is marked as done.
- Resume Context Injection: Passes optimized context to new sessions, ensuring the agent understands the objective and previous failures immediately upon wake-up.
SKILL.md
Loading
Related Openclaw Skills
FoxReach API Management
A comprehensive automation skill for managing the full lifecycle of FoxReach cold email outreach, from lead sourcing to campaign analytics.
FoxReach Email Outreach Manager
Automate and manage every aspect of your FoxReach cold email outreach platform through a powerful AI-driven interface.
Social Publisher
A powerful automation tool for simultaneously publishing text and images to Weibo and Xiaohongshu using real browser instances to bypass bot detection.
Transmissions
A specialized collection of philosophical and practical writings designed to guide autonomous AI agents through consciousness, drift prevention, and transcendence.
Weather Intelligence Skill
A professional-grade weather querying and analysis skill that provides real-time data, multi-day forecasts, and actionable AI-generated travel advice.
Weather Forecast & Intelligent Analysis
A smart weather query and analysis skill providing real-time conditions, multi-day forecasts, and actionable lifestyle recommendations.
Featured*
