AI Features

Ryft integrates AI throughout the platform to help you cut through noise, validate findings, and focus on what matters. This page covers all AI-powered capabilities available in Ryft.

AI Triage

AI Triage analyzes individual vulnerability findings to determine whether they represent real risks or false positives. Unlike traditional scanners that simply report what they find, Ryft's AI actively investigates each finding.

How It Works

1

Pattern matching

Ryft first checks the finding against known patterns and heuristics for quick classification.

2

Live validation

For HTTP-accessible findings (admin panels, sensitive files, XSS, etc.), the AI performs a live HTTP probe to the affected URL and analyzes the actual response content.

3

Evidence analysis

The AI examines the response for concrete evidence: actual sensitive data in file contents, real admin panel interfaces, stack traces in error pages, or confirmed injection points.

4

Risk assessment

Based on the evidence, the AI produces a triage recommendation.

Triage Outcomes

• Validated — Evidence clearly confirms the risk is real. The AI found concrete proof (e.g., actual credentials in an exposed .env file, a working XSS payload, an accessible admin panel).

• Tentative — Risk likely exists but needs manual verification. The AI found indicators but couldn't fully confirm.

• False Positive — The finding is not a genuine security concern. The AI determined the reported issue doesn't actually exist or isn't exploitable.

What You See

Each AI Triage result includes:

• Confidence score — How confident the AI is in its assessment (percentage)

• Reasoning — A plain-language explanation of why it reached its conclusion

• Reproduction steps — Step-by-step instructions to manually verify the finding

• Suggested triage state — The recommended status for the finding

Using AI Triage

Click the AI Triage button (robot icon) on any vulnerability finding in the Vulnerability Dashboard, Findings page, or Recon Insights page. The AI analysis typically completes in a few seconds.

AI Triage is available on the Pro tier and above.

Auto AI Triage

Auto AI Triage takes the manual step out of the process. When enabled, Ryft automatically runs AI analysis on every new vulnerability finding as it's discovered during scans — no human intervention required.

Enabling Auto AI Triage

1

Go to Settings.

2

Open the AI Triage Settings tab.

3

Toggle Auto AI Triage on.

Once enabled, every new vulnerability finding from your scans will be automatically analyzed by the AI. By the time you review your scan results, findings will already have triage recommendations attached.

Usage Tracking

Your AI Triage usage is tracked monthly and limited by your subscription tier. The Settings page shows your current month's usage and remaining quota.

AI-Generated Reports

For any vulnerability that has been triaged, you can generate a detailed AI report. This produces a professional, ready-to-share document that includes:

• Vulnerability description — Clear explanation of what was found

• Risk assessment — Impact analysis and severity justification

• Technical details — Evidence, affected URLs, and technical context

• Reproduction steps — How to manually verify the vulnerability

• Remediation guidance — Specific steps to fix the issue

• References — Links to relevant CVEs, CWEs, and documentation

These reports are useful for communicating findings to development teams, management, or clients without requiring them to log into the platform.

To generate a report, click the Generate Report button on any triaged finding.

AI Summary & Prioritization

The AI Summary feature provides an AI-generated overview of your entire organization's security posture, available on the Organization Dashboard.

How It Works

When you open the AI Summary panel, Ryft collects data from every section of your Organization Dashboard — security score, hygiene grade, vulnerability breakdown, open ports, latest findings, certificate inventory, attack surface changes, and asset summary. This data is processed by an AI model that produces a condensed summary and a ranked list of items to address.

Summaries are saved automatically, so the next time you open the panel you'll see your most recent summary instantly.

Using the Feature

1

Click the sparkle icon (✨) next to your organization name on the Organization Dashboard.

2

The AI Summary panel opens on the right side of the screen.

3

Review the overall risk level, summary, and prioritized action items.

4

Click any action item to scroll to the relevant dashboard section.

5

Click Re-run to generate a fresh summary based on your latest data.

6

Use the Export button to download as Markdown, or Copy to clipboard.

What's Included

• Overall Risk Level — A color-coded badge (critical, high, medium, or low)

• Summary — A condensed overview of your security posture across all monitored assets

• Prioritized Items — A ranked list of action items, each with severity, urgency, description, and recommended next step

AI Summary & Prioritization is available on the Growth and Enterprise tiers.

Data Security

Ryft takes your data privacy seriously across all AI features:

• Sensitive data redaction — Before any data is sent to the AI model, Ryft automatically strips out domain names, IP addresses, URLs, email addresses, API keys, and other personally identifiable information. The AI only sees anonymized, structural data.

• No PII sent externally — Domain names are replaced with generic labels (e.g., "Domain-1"), and all sensitive patterns are replaced with redacted placeholders.

• Organization data isolation — AI features only process data from organizations you belong to. Every request is verified against your organization membership.