Files
Anthropic-Cybersecurity-Skills/skills/detecting-fileless-attacks-on-endpoints/scripts/process.py
T

56 lines
2.2 KiB
Python

#!/usr/bin/env python3
"""Fileless Attack Detector - Scans PowerShell logs for fileless attack indicators."""
import json, csv, re, sys, os
from collections import Counter
from datetime import datetime
FILELESS_PATTERNS = {
"encoded_command": r"(?i)(-enc\s|-e\s|-encodedcommand|frombase64string)",
"download_cradle": r"(?i)(downloadstring|invoke-webrequest|net\.webclient|wget\s|curl\s)",
"amsi_bypass": r"(?i)(amsiutils|amsiinitfailed|amsi\.dll)",
"reflection": r"(?i)(system\.reflection|loadassembly|gettype.*invoke)",
"wmi_abuse": r"(?i)(win32_process.*create|wmiclass|managementclass)",
"credential_access": r"(?i)(mimikatz|invoke-mimikatz|sekurlsa|logonpasswords)",
"invoke_expression": r"(?i)(iex\s|invoke-expression)",
}
def scan_powershell_logs(csv_path: str) -> list:
detections = []
with open(csv_path, "r", encoding="utf-8-sig") as f:
for row in csv.DictReader(f):
script = row.get("ScriptBlockText", row.get("Message", ""))
for pattern_name, pattern in FILELESS_PATTERNS.items():
if re.search(pattern, script):
detections.append({
"timestamp": row.get("TimeCreated", row.get("Date and Time", "")),
"host": row.get("Computer", row.get("MachineName", "")),
"technique": pattern_name,
"script_excerpt": script[:300],
})
break
return detections
def generate_report(detections: list, output_path: str) -> None:
by_technique = Counter(d["technique"] for d in detections)
report = {
"report_generated": datetime.utcnow().isoformat() + "Z",
"total_detections": len(detections),
"by_technique": dict(by_technique),
"detections": detections[:100],
}
with open(output_path, "w") as f:
json.dump(report, f, indent=2)
if __name__ == "__main__":
if len(sys.argv) < 2:
print("Usage: python process.py <powershell_logs.csv>")
sys.exit(1)
detections = scan_powershell_logs(sys.argv[1])
out = os.path.join(os.path.dirname(sys.argv[1]) or ".", "fileless_detection_report.json")
generate_report(detections, out)
print(f"Detections: {len(detections)}")