Datasets:
|
Download README.md from logtriage/security-log-samples: direct link, hf CLI and curl.
- Browser
- Download file 2.13 kB
-
https://huggingface.co/datasets/logtriage/security-log-samples/resolve/main/README.md
- Command line
-
hf download hf://datasets/logtriage/security-log-samples/README.md
-
curl -L -o README.md https://huggingface.co/datasets/logtriage/security-log-samples/resolve/main/README.md
2.13 kB
| license: cc-by-4.0 | |
| pretty_name: Security Log Samples (malicious and benign, 50+ formats) | |
| language: | |
| - en | |
| tags: | |
| - cybersecurity | |
| - security | |
| - logs | |
| - log-analysis | |
| - threat-detection | |
| - detection-engineering | |
| - sigma | |
| - siem | |
| task_categories: | |
| - text-classification | |
| size_categories: | |
| - n<1K | |
| configs: | |
| - config_name: manifest | |
| data_files: manifest.csv | |
| # Security log samples | |
| Paired **malicious** and **benign** sample logs for more than 50 security log formats, for testing log | |
| parsers, detection rules and analytics, plus the monthly data behind two LogTriage research reports. | |
| Published by [LogTriage](https://logtriage.app); the same files are on | |
| [GitHub](https://github.com/logtriage-cyber/security-log-samples). Licensed CC BY 4.0. | |
| Each attack sample has a benign counterpart from the same source, so a rule can be checked both ways: it | |
| must fire on the first and stay silent on the second. Public IP addresses are replaced with documentation | |
| ranges (RFC 5737). `manifest.csv` (shown in the viewer above) lists every file with its format, label | |
| (`malicious` / `benign`), size, line count, SHA-256 and the | |
| [LogTriage Sigma rules](https://github.com/logtriage-cyber/logtriage-detection-rules) validated against it. | |
| ## Monthly research data | |
| | Path | What it is | Report | | |
| |---|---|---| | |
| | `reports/server-attacks/<YYYY-MM>.json` | What attacked an ordinary internet-facing server that month: SSH and web attack volumes, the most requested probe paths, scanner user agents, attacking networks and countries. Aggregates only, never an IP address. | [What attacks an internet-facing server](https://logtriage.app/reports/server-attacks/) | | |
| | `reports/ip-blocklists/<YYYY-MM>.json` | How much of that month's real attack traffic each of 17 free IP blocklists would have caught, alone and in combination. Aggregates only. | [Which free IP blocklists work](https://logtriage.app/reports/ip-blocklists/) | | |
| Field-by-field guides to the formats are in the [log format reference](https://logtriage.app/reference/log-formats/); | |
| the [free log analyzer](https://logtriage.app/tools/log-analyzer/) reads every format here in the browser. | |