cfn-error-management

The cfn-error-management mega-skill serves as the centralized backbone for reliability within the Claude Flow Novice (CFN) framework. It consolidates error handling, batch processing, and persistent logging, enabling developers and AI agents to manage complex execution failures systematically. By providing a unified interface for tracking agent-specific errors—such as timeouts, process crashes, validation failures, and coordination mismatches—this tool ensures that errors are not just identified but categorized and prepared for intelligent recovery or batch processing. It is designed for engineers and AI orchestrators who require robust, observable, and reproducible error workflows in production environments. Whether dealing with individual agent failures during development or large-scale batch processing tasks, this skill ensures that error data remains structured, searchable, and actionable throughout the agent's lifecycle.

• Standardized Error Capture: Provides consistent schemas for capturing agent-specific error types (TIMEOUT, CRASH, VALIDATION, COORDINATION, UNKNOWN), ensuring context and error metadata are preserved.

• Intelligent Batch Processing: Allows developers to group multiple errors for efficient analysis, wave-based parallel processing, and automated remediation via pre-defined batch templates.

• Structured Logging and Cleanup: Implements structured log storage with built-in rotation and cleanup utilities to prevent log bloat while maintaining a clear audit trail for debugging.

• TypeScript and Bash Integration: Offers flexible CLI wrappers (Bash) for quick operational tasks and TypeScript-based core modules for deep integration into agent orchestration logic.

• Usage Notes: Best utilized as part of the CFN Loop validation pipeline to capture errors from Loop 3 agents and pass them to subsequent review or recovery logic.

• Prerequisites: Requires Bash 4.0+ and jq for CLI operations; Node.js is recommended for advanced TypeScript-based logging features.

• Inputs: Accepts standard agent identification, task metadata, and specific error codes for ingestion into the system.

• Constraints: Designed specifically for the CFN Loop environment; while modular, it relies on standard directory structures defined within the repository's skill architecture.