SOC Detection & Incident Analysis Platform
SentinelLog analyzes security log files offline, raises MITRE ATT&CK-mapped alerts, correlates them into attack chains, scores the overall risk and writes incident reports. It is a learning/portfolio project: pure Python, no database, no web framework, no cloud services.
- 20 log formats parsed (SSH, Windows Security, Sysmon, Firewall, DNS, Apache, Nginx, IIS, Active Directory, VPN, Cloud and more)
- 27 detection rules with MITRE ATT&CK mapping (28 alert types including
ATTACK_CHAIN) - Attack Chain Correlation Engine
- Threat Scoring (0-100 risk score)
- Extended IOC extraction (12 categories)
- 4 report formats: TXT, JSON, HTML Dashboard, PDF
- Deduplication
- Alert IDs (
ALERT-0001, ...) - Enterprise support: Kerberos (Kerberoasting, AS-REP Roasting, S4U), VPN (Brute Force, Impossible Travel), Cloud (AWS/Azure/GCP persistence)
pip install -r requirements.txt
python src/cli.py analyze-all sample_logs/reports/incident_report.txt - Human-readable
reports/incident_report.json - Machine-readable
reports/incident_report.html - Interactive dashboard
reports/incident_report.pdf - SOC exportpython src/cli.py analyze-all sample_logs/ --output reports/ --window 10
python src/cli.py analyze single.log
python src/cli.py version
python src/cli.py rulesOptions for analyze and analyze-all:
| Option | Default | Meaning |
|---|---|---|
--output DIR |
reports |
Output directory for the four reports |
--window MIN |
10 |
Attack-chain correlation window in minutes |
--no-dedup |
off | Keep repeated (source_ip, alert_type) alerts |
| Alert Type | Severity | MITRE ID | Description | Log Source |
|---|---|---|---|---|
AUTH_BRUTE_FORCE |
MEDIUM-CRITICAL* | T1110 | Repeated failed logins from one IP within 5 minutes (4+ failures) | SSH, Windows Security |
AUTH_FAILURES_THEN_SUCCESS |
MEDIUM-CRITICAL* | T1110 | Successful login after 4+ failures within 5 minutes | SSH, Windows Security |
MULTI_ACCOUNT_FAILURES |
MEDIUM | T1110 | Failures against 3+ usernames from one IP within 5 minutes | SSH |
ACCOUNT_CREATED |
MEDIUM | T1136.001 | New local account created | Linux useradd, Windows 4720 |
ACCOUNT_DELETED |
MEDIUM | T1531 | Account deleted | Linux userdel, Windows 4726 |
EVENT_LOG_CLEARED |
HIGH | T1070.001 | Windows Event ID 104 (log cleared) | Windows Event Log |
WHOAMI_EXECUTION |
LOW | T1033 | whoami.exe process creation (Event ID 4688) | Windows Security |
POWERSHELL_EXECUTION |
MEDIUM | T1059.001 | powershell.exe process creation (Event ID 4688) | Windows Security |
LARGE_DATA_EXFIL |
HIGH | T1567 | Outbound transfer larger than 25 MB | Network flow |
USB_INSERTION |
LOW | T1200 | USB storage device inserted | Windows device events |
HTTP_BRUTE_FORCE |
MEDIUM | T1110 | 4+ unique HTTP /login failures from one IP within 5 minutes | HTTP access log |
SYSMON_SUSPICIOUS_PROCESS |
HIGH | T1059 | Encoded PowerShell, Mimikatz or rundll32 abuse | Sysmon (Event ID 1) |
PORT_SCAN |
MEDIUM | T1046 | One IP touching 5+ unique ports | Firewall |
DNS_TUNNELING |
HIGH | T1071.004 | 5+ DNS queries with unusually long labels from one IP | DNS |
WEB_RECONNAISSANCE |
MEDIUM | T1595 | One IP probing 3+ admin/sensitive pages | HTTP access log |
DIRECTORY_TRAVERSAL |
HIGH | T1190 | ../../ traversal to sensitive files | Apache, Nginx, IIS |
SQL_INJECTION |
HIGH | T1190 | UNION SELECT in a request | Apache, Nginx, IIS |
COMMAND_INJECTION |
CRITICAL | T1190 | cmd= parameter in a request | Apache, Nginx, IIS |
WEBSHELL |
CRITICAL | T1505.003 | Successful (200) request to a shell-like path | IIS |
KERBEROASTING |
HIGH | T1558.003 | 5+ TGS requests (Event ID 4769) from one IP | Active Directory |
ASREP_ROASTING |
HIGH | T1558.004 | 4+ TGT requests with pre-auth disabled (0x40800000) | Active Directory |
AD_PRIVILEGE_ESCALATION |
CRITICAL | T1098.007 | Member added to a privileged group (Event ID 4732) | Active Directory |
S4U_ABUSE |
HIGH | T1558 | 4+ service-for-user requests (Event ID 4771) from one IP | Active Directory |
VPN_BRUTE_FORCE |
MEDIUM | T1110 | 5+ VPN auth failures from one IP within 5 minutes | OpenVPN, Cisco, Fortinet |
IMPOSSIBLE_TRAVEL |
HIGH | T1078 | Same user authenticating from two IPs within 30 minutes | OpenVPN, Cisco, Fortinet |
CLOUD_BRUTE_FORCE |
MEDIUM | T1110 | 4+ cloud console login failures from one IP within 5 minutes | AWS, Azure, GCP |
CLOUD_PERSISTENCE |
CRITICAL | T1098 | Backdoor user / admin role / access key creation | AWS, Azure, GCP |
ATTACK_CHAIN |
CRITICAL | T1595 | 3+ distinct alert types from one IP within the correlation window | Correlation engine |
* Severity depends on the number of failed attempts. See docs/detections.md.
| Technique | Tactic | Alerts |
|---|---|---|
| T1033 System Owner/User Discovery | Discovery | 1 |
| T1046 Network Service Discovery | Discovery | 1 |
| T1059 Command and Scripting Interpreter | Execution | 1 |
| T1059.001 PowerShell | Execution | 1 |
| T1070.001 Clear Windows Event Logs | Defense Evasion | 1 |
| T1071.004 DNS | Command and Control | 1 |
| T1078 Valid Accounts | Initial Access | 1 |
| T1098 Account Manipulation | Persistence | 1 |
| T1098.007 Account Manipulation: Additional Local or Domain Groups | Persistence | 1 |
| T1110 Brute Force | Credential Access | 6 |
| T1136.001 Create Account: Local Account | Persistence | 1 |
| T1190 Exploit Public-Facing Application | Initial Access | 3 |
| T1200 Hardware Additions | Initial Access | 1 |
| T1505.003 Server Software Component: Web Shell | Persistence | 1 |
| T1531 Account Access Removal | Impact | 1 |
| T1558 Steal or Forge Kerberos Tickets | Credential Access | 1 |
| T1558.003 Steal or Forge Kerberos Tickets: Kerberoasting | Credential Access | 1 |
| T1558.004 Steal or Forge Kerberos Tickets: AS-REP Roasting | Credential Access | 1 |
| T1567 Exfiltration Over Web Service | Exfiltration | 1 |
| T1595 Active Scanning | Reconnaissance | 2 |
Full mapping: docs/mitre_matrix.md.
┌─────────────┐ ┌─────────┐ ┌────────────┐ ┌─────────────┐ ┌──────────┐
│ Log Files │────▶│ Parser │────▶│ Detectors │────▶│ Correlation │────▶│ Reporter │
└─────────────┘ └─────────┘ └────────────┘ └─────────────┘ └──────────┘
│
▼
┌─────────────┐
│ Scoring │
└─────────────┘See docs/architecture.md and docs/attack_chains.md.
python -m pytest tests/ -vSentinelLog/
├── sample_logs/ # 21 sample log files (SSH, Windows, Sysmon, firewall, DNS,
│ # Apache, Nginx, IIS, Active Directory, VPN, cloud, ...)
├── reports/ # generated: incident_report.{txt,json,html,pdf}
├── screenshots/
├── tests/
│ ├── test_apache.py
│ ├── test_dns.py
│ ├── test_http_bruteforce.py
│ ├── test_portscan.py
│ ├── test_sysmon.py
│ ├── test_web_recon.py
│ ├── test_windows_security.py
│ ├── test_correlation.py
│ ├── test_ioc_expansion.py
│ ├── test_scoring.py
│ ├── test_dedup.py
│ ├── test_enterprise.py
│ └── test_end_to_end.py
├── rules/
│ ├── authentication.yaml
│ └── ssh.yaml
├── docs/
│ ├── architecture.md
│ ├── detections.md
│ ├── mitre_matrix.md
│ └── attack_chains.md
├── src/
│ ├── cli.py
│ ├── detector.py
│ ├── parser.py
│ ├── ioc.py
│ ├── mitre.py
│ ├── reporter.py
│ ├── scoring.py
│ ├── severity.py
│ ├── rule_loader.py
│ ├── __init__.py
│ └── test_*.py # original v1.0 unit tests (run with PYTHONPATH=src)
├── requirements.txt
├── README.md
└── .gitignore