10Dataset
The numbers behind
the instrument.
Everything on this page is computed from files in the repository at request time. No warehouse, no nightly job, no cache to go stale.
dataset 2026.10.23seed 2718static generation
27Incidents13 with CVSS
65Stack items10 categories
24Actions478 min total
29Surface linksIncident → component
Severity mix
- critical12
- high13
- medium2
- low0
Vector distribution
- Supply chain5
- Hosting4
- Web server3
- Endpoint3
- Network3
- Data breach2
- State actor2
- Ransomware2
- Certificate authority1
- AI agent1
- Banking1
Effort distribution
- under 10 min5
- 10–20 min12
- 21–40 min6
- over 40 min1
mean 20 min · median 20 min
Coverage
| Surface | Count |
|---|---|
| Incidents with a stack surface | 21 |
| Incidents matched by region only | 6 |
| Certificate records | 5 |
| Playbooks | 8 |
| Actions with a command | 3 |
How to reproduce
bun install bun run --filter @ferroscope/seed test bun run --filter @ferroscope/core test bun run loc
The dataset tests assert exactly 27 incidents, that every referenced stack item and action exists, and that timelines are chronologically ordered. A broken reference fails the build.