How It Works

Every 30 minutes, the pipeline pulls new and updated vulnerability data from each source below, normalizes it into a local SQLite database, computes a priority score for each CVE, and sends Telegram alerts for critical findings. This dashboard is a read-only view into that database.

Data Sources

🏛️

NVD — National Vulnerability Database

NIST's comprehensive CVE catalog. The primary source for vulnerability descriptions, CVSS v3.1/v4.0 scores, severity ratings, CWE weakness classifications, and CPE (affected software/hardware) data.

CVE Details CVSS Scores Severity CWE CPE CVSS Vectors
🛡️

CISA KEV — Known Exploited Vulnerabilities

CISA's authoritative catalog of vulnerabilities confirmed to be actively exploited in the wild. Includes vendor/product identification, remediation due dates, and ransomware-use flags. Any CVE on the KEV list is automatically escalated to ACT_NOW priority.

Active Exploitation Vendor/Product Due Dates Ransomware Flags
📊

FIRST EPSS — Exploit Prediction Scoring System

A probability model from FIRST.org that estimates the likelihood a CVE will be exploited in the next 30 days. Scores range from 0 to 1. EPSS is multiplied by CVSS to produce the dashboard's composite priority score, and drives the Trending page.

Exploit Probability Percentile Rank Trending Detection
📋

MITRE CVE List (cvelistV5)

The upstream CVE assignment authority. Ingested via the CVE Project's GitHub repository API to catch newly assigned CVEs before they appear in NVD, and to fill in CVSS data from CNA (CVE Numbering Authority) submissions.

Early CVE Assignments CNA CVSS Data Descriptions
🔓

OSV / GitHub Security Advisories (GHSA)

GitHub's curated advisory database covering open-source ecosystems — npm, PyPI, Go, RubyGems, Maven, and more. Provides ecosystem-specific severity ratings and maps advisories back to CVE IDs for correlation with other sources.

Open Source CVEs Ecosystem Coverage Advisory Severity
🔬

nomi-sec/PoC-in-GitHub

An enrichment source that tracks proof-of-concept exploit repositories on GitHub. The pipeline reads metadata only — repo URL and star count — and never clones or executes any code. PoC availability is a key factor in the Hunt List prioritization.

PoC URLs Star Counts Hunt List Signals

Priority Scoring

Each CVE receives a composite priority score and is assigned to one of four tiers:

ACT NOW On the KEV list, flagged for ransomware use, or priority score exceeds threshold
HIGH Priority ≥ 2.0, or critical/high severity with EPSS ≥ 0.3
MEDIUM Medium or high severity CVEs not meeting HIGH criteria
INFO Low severity or insufficient data for higher classification

Priority score = EPSS × CVSS. This weights real-world exploit likelihood against technical severity, surfacing CVEs that are both dangerous and likely to be exploited soon.

Alerting

The pipeline sends three types of Telegram notifications: push alerts for ACT_NOW tier events as they're detected, batch reports consolidating all new events from an ingest cycle, and a daily digest summarizing the last 24 hours of activity across all tiers.

Infrastructure

Self-hosted on a single VPS. The pipeline is Python, the database is SQLite with WAL mode, and this dashboard is Flask served behind nginx with TLS. The dashboard connection is read-only — it cannot modify the database. All source code is local, not on GitHub.