What problem does it solve?
This Skill addresses the need for comprehensive, end-to-end stability audits of the agent-harness system, ensuring its operational stability across various components.
Core Features & Use Cases
- Exhaustive Audit: Perform a thorough examination of agent-harness stability, covering installation, hooks, MCP, daemon, state, and process hygiene.
- Safety Model: Implement a safety-first approach, ensuring minimal impact on system state and processes.
- Quick Path: Offers a streamlined process to quickly verify the stability of agent-harness without installation or cleanup.
- Manual Workflow: Detailed manual steps for advanced users seeking a more granular audit.
- Completion Report: Generates a comprehensive report outlining the audit's findings and actions taken.
Quick Start
Run a quick stability audit with the command: python3 skills/stability-audit/scripts/e2e_stability_audit.py --json