Security Monitoring Setup
SIEM configurations, log correlation rules, threat detection queries, and security dashboard templates.
Product Preview
Try the interactive reader and demo tools below, or get the full product with all content unlocked.
Interactive Reader (Free Preview) ⚙ Try Demo Tools Download Free SampleFile Structure 34 files
Documentation Preview README excerpt
Security Monitoring Setup
A complete, code-forward blueprint for a Linux/cloud detection and response
pipeline — from shipping raw logs all the way to ranked, correlated incidents.
This is not a vendor brochure: it is the actual config, detection rules, and a
dependency-free correlation engine you deploy and adapt.
The stack: Filebeat → Logstash (ECS normalization) → Elasticsearch, with
Sigma detections, Falco runtime rules, auditd syscall auditing,
Prometheus + ElastAlert alerting, a Grafana dashboard, threat-hunting
queries, a MITRE ATT&CK coverage map, and an incident-triage runbook — tied
together by scripts/log_triage.py, a stdlib-only correlation tool that turns a
stream of findings into scored incidents.
Every file is annotated with the why. All addresses use RFC1918 internal ranges
and RFC5737 documentation ranges; no real hosts, keys, or credentials appear
anywhere.
Who this is for
Security engineers, SREs, and platform teams standing up (or sharpening) a
detection capability without buying a six-figure SIEM. You should be comfortable
on Linux and with JSON/YAML; the detections assume basic familiarity with the
MITRE ATT&CK vocabulary, which mitre/attack-mapping.md reinforces.
What you get
- A real ingestion pipeline: annotated
filebeat.ymland a Logstash pipeline
that parses auth/sudo logs and normalizes everything to **Elastic Common
Schema** so detections are written once against stable field names.
- High-signal detections: five Sigma rules (SSH brute force→success, web
shell, curl|bash, sudoers tampering, AWS root login) and six Falco
runtime rules, each tagged to MITRE ATT&CK.
- Syscall auditing: a curated
auditd/audit.ruleswatching the files,
binaries, and syscalls that matter — without drowning you in noise.
- Correlation that thinks in attack chains: a documented correlation model
and scripts/log_triage.py, a stdlib-only engine that detects four named
attack chains, scores them, and dedups — its output is the spec.
- Alerting + visualization: Prometheus rules for pipeline health and signal
spikes, an ElastAlert rule for live brute-force alerting, and an importable
Grafana Security Overview dashboard.
- Hunting + response: eight threat-hunting queries (KQL/DSL/SPL), a MITRE
ATT&CK coverage matrix with an honest gap analysis, and a battle-tested
incident-triage runbook.
Prerequisites
| Tool | Why | Notes |
|---|---|---|
| Filebeat ≥ 8 | log shipping | filebeat test config validates the bundled config |
| Logstash ≥ 8 | parsing + ECS normalization | pipeline is portable to Vector/Fluentd |
| Elasticsearch/OpenSearch ≥ 8 | store + search | dashboard + hunts target security-* indices |
| Falco ≥ 0.36 | runtime/syscall detection | falco -L validates the local rules |
| auditd | syscall auditing | auditctl -R audit.rules to load |
| Prometheus + Grafana | alerting + dashboards | promtool check rules validates the alerts |
| Python ≥ 3.10 | log_triage.py | stdlib only, no pip install |
... continues with setup instructions, usage examples, and more.