How to use AI Policy & Compliance Narration in rConfig V8
How to use AI Policy & Compliance Narration in rConfig V8
Section titled “How to use AI Policy & Compliance Narration in rConfig V8”This capability does two jobs. Narrate with AI turns a compliance result, with its pass and fail rule lines, into a short report you can paste into an audit response or hand to the device owner. AI Generate Rules drafts a new policy definition for you, either generic CIS/NIST-style rules or rules based on a config you paste in, so you start from a working draft instead of a blank editor.
Prerequisites
Section titled “Prerequisites”- The AI Center set up with an active provider (see Getting Started).
- Policy & Compliance Narration turned on in AI Configuration. This one toggle controls both narration and rule generation.
- You’ve accepted the AI Center terms. You need permission to view compliance results for narration, and to create or edit policy definitions for rule generation.
Narrate a compliance result
Section titled “Narrate a compliance result”1. Open the result
Section titled “1. Open the result”Go to Compliance, open a report, and click a device’s result row. The result detail page opens. You can also get there from a device’s compliance panel.
2. Click Narrate with AI
Section titled “2. Click Narrate with AI”In the Policy Method Results header, next to Show JSON, click Narrate with AI. It reads Analyzing… while the request runs.
3. Read the narration
Section titled “3. Read the narration”An AI Compliance Narration card appears with four sections:
| Section | What it tells you |
|---|---|
| Summary | The overall result in a sentence or two |
| Failed Rules Analysis | Each failed rule in plain English: what it checks and why the device failed |
| Security Impact | Why the failures matter |
| Recommended Remediation | What to change on the device to pass |
Remediation is advice for you to apply through your normal change process. rConfig doesn’t push anything to the device.
Narration sends the device name, policy name, overall result, the reason, and the per-rule results. It never sends configuration text.
Draft policy rules with AI
Section titled “Draft policy rules with AI”1. Open the policy definition editor
Section titled “1. Open the policy definition editor”Go to Policy Definitions and open the definition you’re drafting in the policy definition editor. Give it a name first; rule generation needs one.
2. Open AI Generate Rules
Section titled “2. Open AI Generate Rules”Click the AI Generate Rules icon in the editor toolbar. A panel opens: Paste device config to generate device-specific rules, or leave blank for generic CIS/NIST rules.
3. Choose generic or device-specific rules
Section titled “3. Choose generic or device-specific rules”| Leave Device Config (optional)… | You get |
|---|---|
| Empty | Generic CIS/NIST-style rules for the platform implied by your policy name and description |
| Filled with a running config | Rules tailored to that config, for example matching its actual interface and logging setup |
Credentials in the pasted config are redacted before sending, and up to 8 KB of it is used.
4. Generate and confirm
Section titled “4. Generate and confirm”Click Generate Rules (it reads Generating… while running). rConfig then asks Replace editor content? because the draft replaces whatever is in the editor. Click Replace to accept it.
5. Review, edit, and save
Section titled “5. Review, edit, and save”The rules land in the editor as an unsaved draft. Read every rule, edit or delete what doesn’t fit your standard, then save as normal. Nothing reaches the policy store until you save.
Why this is the one capability that writes
Section titled “Why this is the one capability that writes”Every other AI feature only reads. Rule generation produces text that you can save to rConfig’s own policy store. It still never touches a device, and it writes nothing until you click save.