Afrexai Postmortem

What This Skill Does

The Afrexai Postmortem skill is a specialized documentation engine designed to transform chaotic, real-time incident data into professional, SRE-standard postmortem reports. By processing raw inputs—such as copy-pasted Slack threads, hurried notes, or rough timelines—this skill leverages structured logic to synthesize incidents into a coherent, blameless format. It is designed to save engineering teams time while ensuring that critical learnings from outages are captured, analyzed, and transformed into actionable prevention measures. It effectively bridges the gap between the frantic nature of incident response and the formal requirement for historical record-keeping.

Installation

To integrate this tool into your OpenClaw workspace, execute the following command in your CLI: clawhub install openclaw/skills/skills/1kalin/afrexai-postmortem

Use Cases

This skill is essential for DevOps, SRE, and engineering teams maintaining production environments. It is best utilized immediately following the resolution of any service disruption (SEV1 through SEV4). Common use cases include:

1. Standardizing Incident Reporting: Ensuring every incident, regardless of scope, follows the same format for historical tracking.
2. Blameless Culture Enforcement: Automatically rephrasing incident descriptions to focus on systemic failures rather than individual human error.
3. Post-Incident Remediation: Generating high-quality, prioritized task lists for Jira or Trello to address underlying technical debt.
4. Compliance Documentation: Creating auditable records for uptime SLA monitoring and SOC2 reporting requirements.

Example Prompts

1. "I need a postmortem for this incident: We had a DB connection pool exhaustion at 03:00 UTC due to a bad query. It affected 40% of users for 50 minutes. We fixed it by increasing the max connection limit and reverting the last change. Can you format this?"
2. "Here is the Slack export from our outage this morning: [PASTE DATA]. Please generate a SEV2 postmortem and suggest three long-term architectural improvements."
3. "Help me finalize this draft. The timeline is: Detected 10:00, Investigated 10:15, Identified culprit 10:30, Resolved 10:45. What are some preventative measures for a memory leak like this?"

Tips & Limitations

To get the best output, provide as much context as possible during the initial prompt. While the tool is excellent at structuring data, it relies on the accuracy of the provided notes; therefore, always verify the timestamps and impact statistics before finalizing your report. Remember that the skill thrives on the '5 Whys' method—if you provide only a single reason, it may struggle to find the deeper systemic cause. For the highest ROI, use this for all incidents rated SEV3 and above to begin identifying recurring patterns across your infrastructure.