AI Narrative [Command+]
The AI Narrative is a daily briefing generated by a large language model that analyzes your current posture data, risk register, compliance gaps, and integration feeds to produce a concise, plain-language summary of your organization's top three security exposure areas and recommended actions.
What It Does
Each morning, CMDash's AI pipeline processes the previous 24 hours of data and generates a Narrative with the following structure:
- Exposure Area 1, 2, and 3 — the three most significant areas of risk or compliance gap, each described in plain language
- Supporting data — specific metrics, control gaps, or vulnerability counts that substantiate each exposure area
- Recommended actions — concrete steps you can take today to address each exposure area, with links to the relevant CMDash section (Risk Register entry, Remediation task, Compliance control, or Integration) where available
The Narrative is designed to be readable in under 3 minutes and usable directly in a morning standup or leadership briefing.
When It Refreshes
The Narrative refreshes every morning. The refresh time is early morning in your workspace's configured timezone (default UTC). When the new Narrative is ready, you receive an in-app notification and (if configured) an email or Slack message.
On the Dashboard, a banner indicates when the current Narrative was generated. If you access the Dashboard before the morning refresh completes, the previous day's Narrative is displayed with a "Generated yesterday" label.
How to Interpret Recommendations
The Narrative's recommendations are generated from your actual data, but they are a starting point for human judgment, not a directive. Treat each recommendation as:
- A signal worth investigating — click the linked resource to examine the underlying data
- A candidate for a remediation task — if the recommendation identifies a specific gap, create a task in the Remediations section with the recommended action as the task description
- Input for prioritization — the Narrative ranks exposure areas by the LLM's assessment of relative severity; use this alongside your own context when deciding where to focus
The Narrative does not take actions on your behalf. No changes to risk statuses, control statuses, or remediation tasks are made automatically.
Providing Feedback on Narrative Quality
At the bottom of each Narrative, two buttons allow you to rate the quality: thumbs up (useful and accurate) or thumbs down (inaccurate, irrelevant, or unhelpful). If you select thumbs down, a short optional text field allows you to describe the issue.
Feedback is used to improve the Narrative model over time. It is reviewed by CMDash's AI team and does not contain any of your organization's security data.