Security, Science & HardwareDone with n8n
Monitor and score CVEs with real-time alerts
An n8n workflow monitors CVEs from sources like CISA and OTX, scores them with AI, and alerts via Slack, Gmail and Google Sheets.
- What They Did
- Bright Data nodes in n8n scrape CISA, NIST, CERT-FR and BleepingComputer. A Cohere Command-R agent with session memory extracts CVE details and scores severity, using CISA KEV, CIRCL and AlienVault OTX enrichment. Output parsing and normalization stop it recalculating scores arbitrarily. Hash-based diff checks remove duplicates. Alerts go to Slack, Gmail and Google Sheets, with retries on source failures.
- How It Went
- The author reports 100+ CVEs a day, alerts in under 5 minutes, and 99.8% uptime. The Slack Ack/Investigate/Dismiss buttons are only a demo, because the webhook was blocked, so the system is about 90% production-ready.
- Worth Knowing
- Without strict output constraints and parsing, the AI scored inconsistently. The uptime and false-positive figures are the author's own claims, not independently verified.
Try it yourself with n8n
Build an n8n workflow that checks CVE feeds such as CISA and OTX every [interval], asks an AI model to score each new CVE for relevance to my stack [list of technologies] and severity, and logs them to my Google Sheet [sheet name]. High-scoring items should also alert me in [Slack channel or Gmail]. You're done when a test run logs new CVEs and flags the relevant ones; show me a sample alert before activating sending.