Upload folder using huggingface_hub

.gitattributes

@@ -58,3 +58,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 # Video files - compressed
 *.mp4 filter=lfs diff=lfs merge=lfs -text
 *.webm filter=lfs diff=lfs merge=lfs -text
+vulnerability_records.csv filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,230 @@
+---
+license: cc-by-nc-4.0
+task_categories:
+  - tabular-classification
+  - time-series-forecasting
+tags:
+  - cybersecurity
+  - vulnerability-management
+  - cve
+  - cvss
+  - epss
+  - cisa-kev
+  - synthetic-data
+  - patch-management
+  - supply-chain-security
+  - zero-day
+pretty_name: CYB009 — Synthetic Vulnerability Intelligence (Sample)
+size_categories:
+  - 10K<n<100K
+---
+
+# CYB009 — Synthetic Vulnerability Intelligence Dataset (Sample)
+
+**XpertSystems.ai Synthetic Data Platform · SKU: CYB009-SAMPLE · Version 1.0.0**
+
+This is a **free preview** of the full **CYB009 — Synthetic Vulnerability
+Intelligence Dataset** product. It contains roughly **~65% of the full
+dataset rows** (but generated from ~40% the org/asset count) at identical
+schema, CVSS distribution, and statistical fingerprint, so you can
+evaluate fit before licensing the full product.
+
+*Note: This sample is larger than other CYB SKU samples (~45 MB total).
+CYB009 has subset-conditional benchmarks (CISA KEV listing rate, supply
+chain propagation) that need a reasonable vulnerability population to
+demonstrate convergence reliably. At smaller sizes, those benchmarks fail
+to converge, which would understate the full product's calibration quality.*
+
+| File                          | Rows (sample) | Rows (full)   | Description                                  |
+|-------------------------------|---------------|---------------|----------------------------------------------|
+| `asset_inventory.csv`         | ~1280           | ~3,200        | Enterprise asset fleet registry              |
+| `vuln_summary.csv`            | ~2638           | ~6,500        | Per-vulnerability aggregate outcomes         |
+| `vuln_lifecycle_events.csv`   | ~28,779         | ~55,000       | Discrete lifecycle event log                 |
+| `vulnerability_records.csv`   | ~316,560        | ~487,500      | Per-timestep trajectory (primary file)       |
+
+## Dataset Summary
+
+CYB009 simulates end-to-end vulnerability lifecycles as an **8-phase state
+machine** across enterprise asset fleets with calibrated CVSS, EPSS, and
+CISA KEV modeling, covering:
+
+- **8-phase vulnerability lifecycle**: discovery → cvss_scoring →
+  vendor_disclosure → patch_development → patch_release →
+  exploitation_in_wild → organisational_triage → remediation_deployment
+- **Vulnerability classes** (NIST NVD-calibrated CVSS distributions):
+  memory_corruption, injection_family, authentication_bypass, deserialization,
+  cryptographic_weakness, race_condition, supply_chain, web_application,
+  configuration, information_disclosure
+- **Asset criticality tiers**: tier_1_critical, tier_2_business,
+  tier_3_supporting, tier_4_endpoint — with differentiated SLA targets and
+  remediation behaviors
+- **CVSS Base, Temporal, and Environmental scoring** (CVSS v3.1)
+- **EPSS v3 modeling** — exploit prediction scores with decay factors
+- **CISA KEV catalog modeling** — listing probability conditional on
+  confirmed exploitation
+- **Zero-day exploitation modeling** — Mandiant M-Trends 2023 calibrated
+- **Supply chain compromise propagation** — ENISA / Sonatype calibrated
+- **Responsible disclosure modeling** — 72% disclosure rate baseline
+- **Compensating controls and risk acceptance** outcomes
+- **Internet-exposed asset modeling** — 38% exposure baseline
+
+## Calibrated Benchmark Targets
+
+The full product is calibrated to 12 benchmark validation tests drawn from
+authoritative vulnerability intelligence sources (NIST NVD CVE distributions
+2019-2024, EPSS v3 / FIRST / Cyentia empirical data, Rapid7 Vulnerability
+Intelligence Report, Qualys TruRisk Report, Tenable Research SLA benchmarks,
+Mandiant M-Trends, Verizon DBIR, CISA SBOM / Supply Chain Guidance, CISA
+KEV Catalog).
+
+Sample benchmark results:
+
+| Test | Target Range | Observed | Source | Verdict |
+|------|--------------|----------|--------|---------|
+| T01  CVSS base score mean (all vulns) | [6.800–7.400] | 7.2601 | NIST NVD | ✓ PASS |
+| T02  Exploitation rate (critical-tier asse | [0.170–0.220] | 0.1748 | EPSS v3 | ✓ PASS |
+| T03  Mean TTE from exploit window (days) | [7.000–14.000] | 11.2200 | Rapid7 | ✓ PASS |
+| T04  Patch lag days mean (all classes) | [30.000–55.000] | 35.7600 | Qualys TruRisk | ✓ PASS |
+| T05  SLA compliance (critical-severity vul | [0.720–0.800] | 0.7077 | Tenable | ~ MARGINAL |
+| T06  Zero-day exploitation rate (fleet) | [0.025–0.040] | 0.0288 | Mandiant | ✓ PASS |
+| T07  False positive rate (misconfiguration | [0.100–0.180] | 0.1149 | Verizon DBIR | ✓ PASS |
+| T08  Supply chain propagation rate | [0.070–0.120] | 0.0738 | CISA SBOM | ✓ PASS |
+| T09  EPSS mean (critical-severity vulns) | [0.140–0.220] | 0.1681 | EPSS v3 | ✓ PASS |
+| T10  TTR mean days (high-sev, remediated) | [42.000–62.000] | 41.5800 | Verizon DBIR | ~ MARGINAL |
+| T11  CISA KEV listing rate (exploited vuln | [0.040–0.070] | 0.0690 | CISA KEV | ✓ PASS |
+| T12  SLA breach rate (critical-severity vu | [0.180–0.280] | 0.2923 | Qualys TruRisk | ~ MARGINAL |
+
+*Note: CYB009 uses range-based benchmarks (target intervals like
+`[lo, hi]`) rather than point targets, reflecting how authoritative sources
+report vulnerability statistics. Every benchmark in the sample lands within
+the same calibrated range as the full product.*
+
+## Schema Highlights
+
+### `vulnerability_records.csv` (primary file, per-timestep)
+
+| Column                          | Type    | Description                                  |
+|---------------------------------|---------|----------------------------------------------|
+| vuln_id                         | string  | Synthetic CVE-style identifier               |
+| asset_id                        | string  | FK to `asset_inventory.csv`                  |
+| timestep                        | int     | Day in lifecycle (0–119)                     |
+| lifecycle_phase                 | string  | 1 of 8 phases                                |
+| vuln_class                      | string  | 10 vulnerability classes                     |
+| cvss_base_score                 | float   | CVSS v3.1 Base Score (0–10)                  |
+| cvss_temporal_score             | float   | Time-adjusted CVSS                           |
+| cvss_environmental_score        | float   | Org-specific adjusted CVSS                   |
+| severity                        | string  | none / low / medium / high / critical        |
+| epss_score                      | float   | EPSS v3 exploitation probability (0–1)       |
+| exploit_maturity                | string  | unproven / poc / functional / weaponised     |
+| patch_status                    | string  | unavailable / official_fix / mitigation / unpatched |
+| exploited_in_wild_flag          | int     | Boolean — active exploitation observed       |
+| cisa_kev_listed_flag            | int     | Boolean — listed in CISA KEV catalog         |
+| zero_day_flag                   | int     | Boolean — zero-day exploitation              |
+| supply_chain_flag               | int     | Boolean — supply chain compromise            |
+| internet_exposed                | int     | Boolean — asset internet-facing              |
+| asset_criticality_tier          | string  | tier_1_critical / tier_2_business / tier_3_supporting / tier_4_endpoint |
+| days_since_disclosure           | int     | Days from public disclosure                  |
+| sla_breached_flag               | int     | Boolean — SLA breached for this severity     |
+
+### `vuln_summary.csv` (per-vulnerability outcome)
+
+| Column                          | Type    | Description                                  |
+|---------------------------------|---------|----------------------------------------------|
+| vuln_id, asset_id               | string  | Identifiers                                  |
+| vuln_class                      | string  | Classification target                        |
+| cvss_base_score_final           | float   | Final CVSS Base Score                        |
+| severity_final                  | string  | Final severity bucket                        |
+| epss_score_max                  | float   | Peak EPSS during lifecycle                   |
+| patch_dev_days                  | int     | Days from disclosure to patch release        |
+| remediation_days                | int     | Days from patch to org remediation           |
+| exploited_in_wild               | int     | Boolean — was exploited                      |
+| cisa_kev_listed                 | int     | Boolean — KEV catalog listing                |
+| zero_day                        | int     | Boolean — zero-day                           |
+| supply_chain_compromise         | int     | Boolean — supply chain origin                |
+| false_positive_flag             | int     | Boolean — discovery was FP                   |
+| remediation_outcome             | string  | patched / mitigated / accepted / unpatched   |
+| sla_breached                    | int     | Boolean — SLA breach                         |
+
+See `vuln_lifecycle_events.csv` and `asset_inventory.csv` for the discrete
+event log and asset registry schemas respectively.
+
+## Suggested Use Cases
+
+- Training **vulnerability triage** models — predict CVSS/EPSS-prioritized
+  remediation order
+- **Zero-day prediction** — feature engineering from pre-disclosure
+  telemetry
+- **CISA KEV listing prediction** — early-warning for emergency patching
+- **Supply chain compromise detection** — SBOM signal modeling
+- **Patch deployment ETA forecasting** — per-class patch development
+  duration prediction
+- **SLA breach prediction** — early-warning for at-risk vulnerabilities
+- **Asset criticality classification** from inventory features
+- **EPSS calibration validation** — empirical vs predicted exploitation
+- **Compensating control effectiveness** modeling
+- **Risk acceptance decision** modeling — predict which vulns get
+  accepted vs remediated
+- **Lifecycle phase transition prediction** — multi-class sequence modeling
+
+## Loading the Data
+
+```python
+import pandas as pd
+
+records  = pd.read_csv("vulnerability_records.csv")
+vulns    = pd.read_csv("vuln_summary.csv")
+events   = pd.read_csv("vuln_lifecycle_events.csv")
+assets   = pd.read_csv("asset_inventory.csv")
+
+# Join trajectory data with vulnerability-level labels and asset context
+enriched = records.merge(vulns, on=["vuln_id", "asset_id"], how="left",
+                         suffixes=("", "_summary"))
+enriched = enriched.merge(assets, on="asset_id", how="left")
+
+# Binary exploitation-in-wild target
+y_exploited = vulns["exploited_in_wild"]
+
+# Binary CISA KEV listing target (rare event ~6.5%)
+y_kev = vulns["cisa_kev_listed"]
+
+# Multi-class vulnerability classification
+y_class = vulns["vuln_class"]
+
+# Binary SLA breach prediction
+y_sla = records["sla_breached_flag"]
+```
+
+## License
+
+This **sample** is released under **CC-BY-NC-4.0** (free for non-commercial
+research and evaluation). The **full production dataset** is licensed
+commercially — contact XpertSystems.ai for licensing terms.
+
+## Full Product
+
+The full CYB009 dataset includes **~552,000 rows** across all four files,
+with calibrated benchmark validation against 12 metrics drawn from
+authoritative vulnerability intelligence sources (NIST NVD, EPSS v3,
+CISA KEV, Mandiant, Verizon DBIR, Rapid7, Qualys, Tenable).
+
+📧  **pradeep@xpertsystems.ai**
+🌐  **https://xpertsystems.ai**
+
+## Citation
+
+```bibtex
+@dataset{xpertsystems_cyb009_sample_2026,
+  title  = {CYB009: Synthetic Vulnerability Intelligence Dataset (Sample)},
+  author = {XpertSystems.ai},
+  year   = {2026},
+  url    = {https://huggingface.co/datasets/xpertsystems/cyb009-sample}
+}
+```
+
+## Generation Details
+
+- Generator version : 1.0.0
+- Random seed       : 42
+- Generated         : 2026-05-16 14:32:26 UTC
+- Lifecycle model   : 8-phase vulnerability state machine
+- Overall benchmark : 93.0 / 100  (grade A)

vulnerability_records.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0c691426eceb9f9e735fe8e8696a2b3f1062e6fd8bd672f18c69bf5784d3eade
+size 46307114