Meta Debugger is an AI-powered self-healing system that automatically identifies, analyzes, and repairs runtime errors to create resilient autonomous agents.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install meta-debugger
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 meta-debugger 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 Meta Debugger?
Meta Debugger represents a significant leap for autonomous systems within the Openclaw Skills ecosystem. It functions as a meta-layer that monitors execution, providing real-time runtime monitoring and pattern recognition to detect anomalies before they crash a system. By learning from past failures, it builds a sophisticated knowledge base of error patterns, allowing it to transition from simple error reporting to proactive self-healing.
This skill is essential for developers building production-grade AI systems that require high uptime. Instead of manual intervention, Meta Debugger uses stack trace analysis and context tracking to perform root cause analysis, generates potential code patches or configuration fixes, and can even apply them autonomously in safe mode. It effectively closes the loop between error detection and resolution.
Meta Debugger Use Cases
- Automatically recovering from API timeouts or authentication failures in long-running autonomous workflows.
- Building self-healing infrastructure that applies patches to configuration files when drift or errors are detected.
- Reducing developer on-call burden by using Openclaw Skills to handle common runtime exceptions without human intervention.
- Analyzing historical log data to identify recurring architectural bottlenecks and suggesting preventative guardrails.
How Meta Debugger Works
- Real-time Monitoring: The system continuously watches function executions wrapped with the debugger decorator for runtime exceptions or anomalies.
- Root Cause Analysis: Upon failure, it parses the stack trace and evaluates the execution context to understand the exact state of the system.
- Pattern Matching: It compares the current error against a database of known error patterns (e.g., timeout, value_error, connection_error) to find similar past occurrences.
- Fix Generation: The engine generates multiple fix candidates, ranging from simple retries with exponential backoff to complex code patches.
- Autonomous Repair: Depending on configuration, it either suggests the fix to a human or applies it automatically, followed by a validation check to ensure the system is restored to health.
Meta Debugger Setup
To integrate this capability into your project, install the necessary dependencies and initialize the debugger within your Python environment:
pip install json traceback ast
from meta_debugger import MetaDebugger
# Initialize as part of your Openclaw Skills setup
debugger = MetaDebugger(
name="production_assistant",
auto_fix=True,
safe_mode=True
)
Meta Debugger Data Schema & Taxonomy
Meta Debugger organizes its intelligence through structured patterns and strategy schemas to ensure predictable recovery.
Error Pattern Schema
| Field | Description |
|---|---|
| Causes | List of potential root triggers (e.g., network, server_load) |
| Fixes | Recommended recovery actions (e.g., retry, fallback, cache) |
| Prevention | Guardrails to prevent recurrence (e.g., circuit_breaker) |
Recovery Strategies
- Retry Strategy: Includes max_attempts, backoff_base, and max_delay.
- Fallback Strategy: Defines primary and secondary service routes.
- Circuit Breaker: Tracks failure thresholds and timeout periods to prevent cascading failures.
Meta Debugger Advanced Features
- Learning Loop: Automatically records the success rate of every applied fix to improve future suggestions.
- Custom Error Handlers: Define specific logic for domain-specific errors using the
@debugger.error_handlerdecorator. - AI-Powered Guardrails: Dynamically add pre-execution checks to validate inputs and system state before high-risk operations.
- Openclaw Integration: Seamlessly hooks into the Openclaw on_error event for global agent monitoring.
SKILL.md
Loading
Related Openclaw Skills
IPAM/DNS Reconciliation Audit
A comprehensive network audit skill for reconciling IPAM records with live network state and DNS consistency.
US Stock Financials
A powerful financial data extraction tool that pulls comprehensive XBRL data from SEC EDGAR to generate detailed reports and PDF exports.
Market Breadth Analyzer
A powerful data analysis tool that quantifies stock market health by calculating a 0-100 composite score based on six critical breadth components.
Feishu Bitable Architect
A professional business architecture consultant that translates complex workflows into structured, scalable Feishu Bitable solutions.
gstack-agent
A comprehensive software factory suite featuring over 20 specialized AI agents that handle the entire development lifecycle from product ideation to post-deployment monitoring.
DeepSeek Web Query
An automated browser-based skill that leverages DeepSeek's web interface for real-time internet searching and high-fidelity content extraction.
Featured*
