error-guard
System safety and control-plane skill that prevents agent deadlocks and freezes. Provides non-LLM control commands to inspect task state, flush message queues, cancel long-running work, and recover safely without restarting the container. Use when implementing or operating long-running tasks, sub-agents, benchmarks, background monitors (e.g., Moltbook, PNR checks), or when the system becomes unresponsive and needs immediate recovery controls.
Install via CLI (Recommended)
clawhub install openclaw/skills/skills/amar1432/error-guardWhat This Skill Does
The error-guard skill is the foundational safety layer for the OpenClaw AI agent framework. It acts as a dedicated control-plane utility designed to mitigate deadlocks, process freezes, and unrecoverable states during complex agent workflows. Unlike standard skills that rely on LLM reasoning or heavy external I/O, error-guard operates at the architectural level. It provides non-blocking, constant-time commands to monitor system health and force system resets if the agent becomes unresponsive. By managing task registry metadata—such as execution timestamps and state flags—without ever inspecting or storing sensitive message payloads, it offers a secure mechanism to maintain operational stability without requiring container restarts.
Installation
To integrate this safety primitive into your OpenClaw environment, execute the following command in your terminal:
clawhub install openclaw/skills/skills/amar1432/error-guard
Ensure that you have sufficient permissions to modify the active agent's skill registry, as this utility operates as a system-level supervisor.
Use Cases
- Long-Running Workflows: Essential when running complex background tasks, PNR checks, or data synchronization jobs that may hang due to network latency.
- Sub-Agent Orchestration: Ideal for environments managing multiple sub-agents where a single stuck process could impede the parent agent's ability to communicate.
- Automated Benchmarking: Perfect for test environments that require frequent task rotation and automated cleanup to prevent memory bloat or race conditions.
- System Recovery: Acts as an emergency override when the agent becomes unresponsive to standard conversational prompts.
Example Prompts
- "Perform a system status check to identify any stalled tasks or potential deadlocks."
- "Critical failure: flush all active tasks and reset the internal registry immediately."
- "Initiate the recovery sequence to clear hanging background processes and refresh the execution state."
Tips & Limitations
This skill is intended for advanced users and developers who manage high-risk or long-lived workloads. It is not designed for routine interaction. Note that the /flush command is destructive; it will immediately terminate active execution sessions without save-points. Because error-guard is built for speed and reliability, it strictly prohibits model-based reasoning and external polling loops. Always use it as a last line of defense rather than a frequent utility.
Metadata
Not sure this is the right skill?
Describe what you want to build — we'll match you to the best skill from 16,000+ options.
Find the right skillPaste this into your clawhub.json to enable this plugin.
{
"plugins": {
"official-amar1432-error-guard": {
"enabled": true,
"auto_update": true
}
}
}Tags(AI)
Flags: code-execution