Analyze Findings

Analyze Findings

Pull and analyze findings from a completed simulation run or report.

Workflow

1. Identify the report. Ask for the workspace and report ID. If unknown, call list_reports to show recent reports. Alternatively, ask for a run ID and call get_run_status to find the associated report.

2. Fetch findings. Call list_findings with the run ID to get all findings. Call get_report to get the full report with summary data.

3. Analyze and present. Provide:

   • Executive summary — overall quality score and key takeaways
   • Critical issues — findings with critical or high severity that need immediate attention
   • Patterns — common themes across findings (e.g., repeated categories, affected areas)
   • Remediation priorities — ordered list of what to fix first, based on severity and frequency
   • Risk assessment — overall quality and risk posture

4. Deep dive. If the user wants details on specific findings, use list_run_assets to find screenshots, transcripts, or conversation logs that provide context.

Tips

• Group findings by category first (accuracy, security, compliance, UX), then by severity within each category.
• Highlight any finding that appears across multiple conversations — patterns indicate systematic issues.
• If the report shows a low quality score, focus the user on the top 3 highest-impact issues rather than overwhelming them with the full list.