language:
- en
license: cc-by-4.0
task_categories:
- text-classification
- question-answering
tags:
- threat-intelligence
- cve
- vulnerability-management
- risk-management
- nvd
- cisa-kev
- b2b
size_categories:
- 1K<n<10K
CVE Security Intelligence Stream
Enterprise-grade CVE feed for Threat Intelligence and Risk Management teams.
Incremental CVE records from NVD API 2.0 enriched with CISA Known Exploited Vulnerabilities (KEV) metadata. Normalized for LLM SFT/RAG.
B2B Value Proposition
This dataset is built for security vendors, MSSPs, and enterprise SOC/GRC teams who need:
| Use case | What you get |
|---|---|
| Threat Intel enrichment | CVSS severity, affected software, exploit availability per CVE |
| Risk prioritization | CISA KEV flags + CVSS scoring for patch-or-defer decisions |
| Exposure management | Structured affected_software fields mapped from NVD CPE + KEV |
| LLM / RAG pipelines | Clean JSONL rows ready for embedding, summarization, and alert triage |
Integrate in minutes — each record is a self-contained JSON object with stable keys. No scraping, no NVD rate-limit handling, no ETL maintenance on your side.
Live Dataset Statistics
| Metric | Value |
|---|---|
| Total CVE records | 91 |
| Critical + High severity | 54 |
| CISA KEV (actively exploited) | 0 |
| Exploit available / in-the-wild | 1 |
| Average CVSS score | 7.68 |
| Last updated (UTC) | 2026-06-07T19:57:39.233203+00:00 |
Top affected software (sample)
| Software | CVE count |
|---|---|
| Google Chrome | 69 |
| Linux Linux Kernel | 37 |
| Apple Macos | 34 |
| Samsung Android | 8 |
| Microsoft Windows | 6 |
Record Schema (JSONL)
Each line is one valid JSON object:
{
"id": "CVE-2024-1234",
"timestamp": "2026-06-06T12:00:00.000",
"data_value": "Vulnerability description…",
"cvss_score": 9.8,
"cvss_severity": "CRITICAL",
"affected_software": ["Vendor Product"],
"exploit_available": true,
"kev_listed": false,
"cwes": ["CWE-79"],
"published": "…",
"last_modified": "…",
"references": ["https://…"],
"source": "nvd+cisa_kev"
}
Data Sources
Update cadence: every 7–15 minutes (CVE List); pipeline polls every 15 min
Quick Start
from datasets import load_dataset
ds = load_dataset("FXBIA/cve-intelligence-stream", data_files="data/cve_intelligence.jsonl", split="train")
critical = [r for r in ds if r["cvss_severity"] == "CRITICAL"]
exploitable = [r for r in ds if r["exploit_available"]]
print(f"Critical: {len(critical)}, Exploitable: {len(exploitable)}")
Contact & Licensing
Derived from U.S. Government NVD and CISA open feeds.
Maintained by automated Threat Intelligence pipeline (datasetagent).
For commercial integration, API feeds, or private deployment — open an issue on the dataset repo.