Loading
The full pipeline runs every six hours, vulnerability databases are fetched again every hour, and the Exposure Registry refreshes hourly. This page reports what the pipeline did, measured from its own logs. Machine-readable at /api/health.
Hours between the date a source gives an item and the moment the item was stored here, for items stored in the last 30 days. Sources that date items later than they post them are clipped at zero.
| Source | Items | Median | 90th percentile |
|---|---|---|---|
| NVD/CVE Database | 271 | 2.8 h | 4.9 h |
| The Verge (AI) | 128 | 2.7 h | 5.3 h |
| IEEE Xplore (Security & AI Journals) | 123 | 67.7 d | 189.1 d |
| GitHub Advisory Database | 120 | 2.2 h | 4.4 h |
| CNBC Technology | 116 | 2.0 h | 5.0 h |
| The Guardian Technology | 112 | 3.0 h | 6.2 h |
| CSO Online | 69 | 3.6 h | 4.9 h |
| OpenAI Blog | 62 | 11.0 h | 23.9 h |
| Elsevier Security Journals | 60 | 0 min | 0 min |
| The Hacker News | 49 | 4.2 h | 6.9 h |
| SecurityWeek | 47 | 3.1 h | 4.8 h |
| BleepingComputer | 38 | 3.1 h | 5.8 h |
| CISA Known Exploited Vulnerabilities | 36 | 24.0 h | 24.0 h |
| Dark Reading | 32 | 3.2 h | 5.3 h |
| BBC Technology | 30 | 1.9 h | 4.6 h |
| Simon Willison's Weblog | 29 | 3.1 h | 6.8 h |
| MIT Technology Review | 23 | 4.5 h | 6.0 h |
| Schneier on Security | 13 | 1.0 h | 1.8 h |
| Check Point Research | 12 | 5.0 h | 5.0 h |
| Wired (Security) | 11 | 2.0 h | 5.1 h |
| TechCrunch (Security) | 10 | 4.6 h | 5.8 h |
| ACM Digital Library (TOPS, DTRAP, CSUR) | 9 | 0 min | 0 min |
| AWS Security Bulletins | 7 | 4.7 h | 11.7 h |
| AIS eLibrary (Journal of AIS, CAIS, etc.) | 7 | 5.3 d | 23.3 d |
| Microsoft Security Blog | 5 | 2.0 h | 3.7 h |
| AWS Security Blog | 5 | 2.3 h | 2.6 h |
| DeepMind Safety Research | 4 | 3.3 h | 19.2 h |
| CrowdStrike Blog | 3 | 19.0 h | 28.6 h |
0 enabled sources. A source that fails five fetches in a row is paused for 24 hours and then retried. All sources
Every source fetched successfully on its last attempt.
Each pipeline step is recorded as a job. Partial means the step ran out of its time budget and resumes on the next run.
No jobs recorded yet.
How each field is produced: Methods.