Today I'm sharing a new tool: the Containment Failure Report Prompt.
This is a structured prompt designed to generate a containment failure report using a Large Language Model (LLM) after an agent escape. The primary goal is to guide the LLM to produce a comprehensive post-mortem document.
The prompt is broken down into specific sections, directing the LLM to cover all critical areas typically found in such reports. This includes:
- Incident Summary: A brief overview of the event.
- Timeline of Events: A chronological account of the escape.
- Root Cause Analysis: Identification of underlying reasons for the failure.
- Contributing Factors: Other elements that played a role.
- Impact Assessment: Evaluation of the consequences.
- Lessons Learned: Key takeaways from the incident.
- Recommendations: Proposed actions to prevent recurrence.
This tool is for anyone needing to quickly generate a structured and detailed report on an agent containment breach. It helps ensure consistency and completeness in the generated output, reducing the effort required to compile a thorough post-mortem.
Top comments (0)