What it does. Malware IOC extraction is the process of analyzing malicious software Part of mukul975/Anthropic-Cybersecurity-Skills (817 security skills) (mukul975/Anthropic-Cybersecurity-Skills).
Install
npx skills add mukul975/Anthropic-Cybersecurity-Skills --skill performing-malware-ioc-extraction, or copy the skill folder into ~/.claude/skills/performing-malware-ioc-extraction/.- Raw file:
curl -sL https://raw.githubusercontent.com/mukul975/Anthropic-Cybersecurity-Skills/HEAD/skills/performing-malware-ioc-extraction/SKILL.md
SKILL.md (verbatim)
name: performing-malware-ioc-extraction
description: Malware IOC extraction is the process of analyzing malicious software
to identify actionable indicators of compromise including file hashes, network indicators
(C2 domains, IP addresses, URLs), regist
domain: cybersecurity
subdomain: threat-intelligence
tags:
- threat-intelligence
- cti
- ioc
- mitre-attack
- stix
- malware-analysis
- yara
- reverse-engineering
version: '1.0'
author: mahipal
license: Apache-2.0
nist_csf:
- ID.RA-01
- ID.RA-05
- DE.CM-01
- DE.AE-02
mitre_attack:
- T1591
- T1592
- T1593
- T1589
- T1071
Performing Malware IOC Extraction
Overview
Malware IOC extraction is the process of analyzing malicious software to identify actionable indicators of compromise including file hashes, network indicators (C2 domains, IP addresses, URLs), registry modifications, mutex names, embedded strings, and behavioral artifacts. This skill covers static analysis with PE parsing and string extraction, dynamic analysis with sandbox detonation, automated IOC extraction using tools like YARA, and formatting results as STIX 2.1 indicators for sharing.
When to Use
- When conducting security assessments that involve performing malware ioc extraction
- When following incident response procedures for related security events
- When performing scheduled security testing or auditing activities
- When validating security controls through hands-on testing
Prerequisites
- Python 3.9+ with
pefile, yara-python, oletools, stix2 libraries
- Access to malware analysis sandbox (Cuckoo, CAPE, Any.Run, Joe Sandbox)
- VirusTotal API key for enrichment
- Isolated analysis environment (VM or container)
- Understanding of PE file format, common packing techniques
- Familiarity with YARA rule syntax
Key Concepts
Static Analysis IOCs
- File Hashes: MD5, SHA-1, SHA-256 of the sample and any dropped files
- Import Hash (imphash): Hash of imported function table, groups malware families
- Rich Header Hash: PE rich header hash for compiler fingerprinting
- Strings: Embedded URLs, IP addresses, domain names, registry paths, mutex names
- PE Metadata: Compilation timestamp, section names, resources, digital signatures
- Embedded Artifacts: PDB paths, version info, certificate details
Dynamic Analysis IOCs
- Network Activity: DNS queries, HTTP requests, TCP/UDP connections, SSL certificates
- File System: Created/modified/deleted files and directories
- Registry: Created/modified registry keys and values
- Process: Spawned processes, injected processes, service creation
- Behavioral: API calls, mutex creation, scheduled tasks, persistence mechanisms
YARA Rules
YARA is a pattern-matching tool for identifying and classifying malware. Rules consist of strings (text, hex, regex) and conditions that define matching logic. Rules can detect malware families, packers, exploit kits, and specific campaign tools.
Workflow
Step 1: Static Analysis - PE Parsing and Hash Generation
import pefile
import hashlib
import os
def analyze_pe(filepath):
"""Extract IOCs from a PE file through static analysis."""
iocs = {"hashes": {}, "pe_info": {}, "strings": [], "imports": []}
# Calculate file hashes
with open(filepath, "rb") as f:
data = f.read()
iocs["hashes"]["md5"] = hashlib.md5(data).hexdigest()
iocs["hashes"]["sha1"] = hashlib.sha1(data).hexdigest()
iocs["hashes"]["sha256"] = hashlib.sha256(data).hexdigest()
iocs["hashes"]["file_size"] = len(data)
# Parse PE headers
try:
pe = pefile.PE(filepath)
iocs["hashes"]["imphash"] = pe.get_imphash()
iocs["pe_info"]["compilation_time"] = str(pe.FILE_HEADER.TimeDateStamp)
iocs["pe_info"]["machine_type"] = hex(pe.FILE_HEADER.Machine)
iocs["pe_info"]["subsystem"] = pe.OPTIONAL_HEADER.Subsystem
# Extract sections
iocs["pe_info"]["sections"] = []
for section in pe.sections:
iocs["pe_info"]["sections"].append({
"name": section.Name.decode("utf-8", errors="ignore").strip("\x00"),
"virtual_size": section.Misc_VirtualSize,
"raw_size": section.SizeOfRawData,
"entropy": section.get_entropy(),
"md5": section.get_hash_md5(),
})
# Extract imports
if hasattr(pe, "DIRECTORY_ENTRY_IMPORT"):
for entry in pe.DIRECTORY_ENTRY_IMPORT:
dll_name = entry.dll.decode("utf-8", errors="ignore")
functions = [
imp.name.decode("utf-8", errors="ignore")
for imp in entry.imports
if imp.name
]
iocs["imports"].append({"dll": dll_name, "functions": functions})
# Check for suspicious characteristics
iocs["pe_info"]["is_dll"] = pe.is_dll()
iocs["pe_info"]["is_driver"] = pe.is_driver()
iocs["pe_info"]["is_exe"] = pe.is_exe()
# Version info
if hasattr(pe, "VS_VERSIONINFO"):
for entry in pe.FileInfo:
for st in entry:
for item in st.entries.items():
key = item[0].decode("utf-8", errors="ignore")
val = item[1].decode("utf-8", errors="ignore")
iocs["pe_info"][f"version_{key}"] = val
pe.close()
except pefile.PEFormatError as e:
iocs["pe_info"]["error"] = str(e)
return iocs
Step 2: String Extraction and IOC Pattern Matching
import re
def extract_ioc_strings(filepath):
"""Extract IOC-relevant strings from binary file."""
patterns = {
"ipv4": re.compile(
r"\b(?:(?:25[0-5]|2[0-4]\d|[01]?\d\d?)\.){3}"
r"(?:25[0-5]|2[0-4]\d|[01]?\d\d?)\b"
),
"domain": re.compile(
r"\b(?:[a-zA-Z0-9](?:[a-zA-Z0-9-]{0,61}[a-zA-Z0-9])?\.)+"
r"(?:com|net|org|io|ru|cn|tk|xyz|top|info|biz|cc|ws|pw)\b"
),
"url": re.compile(
r"https?://[^\s\"'<>]{5,200}"
),
"email": re.compile(
r"\b[a-zA-Z0-9._%+-]+@[a-zA-Z0-9.-]+\.[a-zA-Z]{2,}\b"
),
"registry": re.compile(
r"(?:HKEY_[A-Z_]+|HKLM|HKCU|HKU|HKCR|HKCC)"
r"\\[\\a-zA-Z0-9_ .{}-]+"
),
"filepath_windows": re.compile(
r"[A-Z]:\\(?:[^\\/:*?\"<>|\r\n]+\\)*[^\\/:*?\"<>|\r\n]+"
),
"mutex": re.compile(
r"(?:Global\\|Local\\)[a-zA-Z0-9_\-{}.]{4,}"
),
"useragent": re.compile(
r"Mozilla/[45]\.0[^\"']{10,200}"
),
"bitcoin": re.compile(
r"\b[13][a-km-zA-HJ-NP-Z1-9]{25,34}\b"
),
"pdb_path": re.compile(
r"[A-Z]:\\[^\"]{5,200}\.pdb"
),
}
with open(filepath, "rb") as f:
data = f.read()
# Extract ASCII strings (min length 4)
ascii_strings = re.findall(rb"[\x20-\x7e]{4,}", data)
# Extract Unicode strings
unicode_strings = re.findall(
rb"(?:[\x20-\x7e]\x00){4,}", data
)
all_strings = [s.decode("ascii", errors="ignore") for s in ascii_strings]
all_strings += [
s.decode("utf-16-le", errors="ignore") for s in unicode_strings
]
extracted = {category: set() for category in patterns}
for string in all_strings:
for category, pattern in patterns.items():
matches = pattern.findall(string)
for match in matches:
extracted[category].add(match)
# Convert sets to sorted lists
return {k: sorted(v) for k, v in extracted.items() if v}
Step 3: YARA Rule Scanning
import yara
def scan_with_yara(filepath, rules_path):
"""Scan file with YARA rules for malware classification."""
rules = yara.compile(filepath=rules_path)
matches = rules.match(filepath)
results = []
for match in matches:
result = {
"rule": match.rule,
"namespace": match.namespace,
"tags": match.tags,
"meta": match.meta,
"strings": [],
}
for offset, identifier, data in match.strings:
result["strings"].append({
"offset": hex(offset),
"identifier": identifier,
"data": data.hex() if len(data) < 100 else data[:100].hex() + "...",
})
results.append(result)
return results
# Example YARA rule for common malware indicators
SAMPLE_YARA_RULE = """
rule Suspicious_Network_Indicators {
meta:
description = "Detects suspicious network-related strings"
author = "CTI Analyst"
severity = "medium"
strings:
$ua1 = "Mozilla/5.0" ascii
$cmd1 = "cmd.exe /c" ascii nocase
$ps1 = "powershell" ascii nocase
$wget = "wget" ascii nocase
$curl = "curl" ascii nocase
$b64 = "base64" ascii nocase
$reg1 = "HKLM\\SOFTWARE\\Microsoft\\Windows\\CurrentVersion\\Run" ascii nocase
condition:
uint16(0) == 0x5A4D and
(2 of ($ua1, $cmd1, $ps1, $wget, $curl, $b64)) or $reg1
}
rule Packed_Binary {
meta:
description = "Detects potentially packed binary"
author = "CTI Analyst"
condition:
uint16(0) == 0x5A4D and
for any section in pe.sections : (
section.entropy >= 7.0
)
}
"""
Step 4: Generate STIX 2.1 Indicators
from stix2 import (
Bundle, Indicator, Malware, Relationship,
File as STIXFile, DomainName, IPv4Address,
ObservedData,
)
from datetime import datetime
def create_stix_bundle(pe_iocs, string_iocs, yara_results, sample_name):
"""Create STIX 2.1 bundle from extracted IOCs."""
objects = []
# Create Malware SDO
malware = Malware(
name=sample_name,
is_family=False,
malware_types=["unknown"],
description=f"Malware sample analyzed: {pe_iocs['hashes']['sha256']}",
allow_custom=True,
)
objects.append(malware)
# File hash indicator
sha256 = pe_iocs["hashes"]["sha256"]
hash_indicator = Indicator(
name=f"Malware hash: {sha256[:16]}...",
pattern=f"[file:hashes.'SHA-256' = '{sha256}']",
pattern_type="stix",
valid_from=datetime.now().strftime("%Y-%m-%dT%H:%M:%SZ"),
indicator_types=["malicious-activity"],
allow_custom=True,
)
objects.append(hash_indicator)
objects.append(Relationship(
relationship_type="indicates",
source_ref=hash_indicator.id,
target_ref=malware.id,
))
# Network indicators from strings
for ip in string_iocs.get("ipv4", []):
if not ip.startswith(("10.", "172.", "192.168.", "127.")):
ip_indicator = Indicator(
name=f"C2 IP: {ip}",
pattern=f"[ipv4-addr:value = '{ip}']",
pattern_type="stix",
valid_from=datetime.now().strftime("%Y-%m-%dT%H:%M:%SZ"),
indicator_types=["malicious-activity"],
allow_custom=True,
)
objects.append(ip_indicator)
objects.append(Relationship(
relationship_type="indicates",
source_ref=ip_indicator.id,
target_ref=malware.id,
))
for domain in string_iocs.get("domain", []):
domain_indicator = Indicator(
name=f"C2 Domain: {domain}",
pattern=f"[domain-name:value = '{domain}']",
pattern_type="stix",
valid_from=datetime.now().strftime("%Y-%m-%dT%H:%M:%SZ"),
indicator_types=["malicious-activity"],
allow_custom=True,
)
objects.append(domain_indicator)
objects.append(Relationship(
relationship_type="indicates",
source_ref=domain_indicator.id,
target_ref=malware.id,
))
bundle = Bundle(objects=objects, allow_custom=True)
return bundle
Validation Criteria
- PE file parsed successfully with hashes, imports, and section analysis
- String extraction identifies network IOCs (IPs, domains, URLs)
- YARA rules match against known malware characteristics
- STIX 2.1 bundle contains valid Indicator and Malware objects
- Private IP ranges and benign strings filtered from IOC output
- IOCs are actionable for blocking and detection rule creation
References
Other files in this skill
assets/template.md (verbatim)
Malware IOC Extraction Report Template
| Field |
Value |
| Filename |
|
| File Size |
|
| File Type |
PE32/PE32+/ELF/Mach-O |
| MD5 |
|
| SHA-1 |
|
| SHA-256 |
|
| Imphash |
|
| SSDeep |
|
| First Seen |
|
PE Analysis
| Attribute |
Value |
| Compile Time |
|
| Entry Point |
|
| Machine Type |
|
| Subsystem |
|
| Is DLL |
|
| Digital Signature |
Valid/Invalid/None |
Sections
| Name |
Virtual Size |
Raw Size |
Entropy |
Suspicious |
| .text |
|
|
|
|
| .data |
|
|
|
|
| .rsrc |
|
|
|
|
Suspicious API Imports
| DLL |
Function |
Purpose |
| kernel32.dll |
VirtualAlloc |
Memory allocation (code injection) |
| kernel32.dll |
CreateRemoteThread |
Process injection |
| wininet.dll |
InternetOpenA |
Network communication |
Network IOCs
IP Addresses
| IP |
Context |
Confidence |
|
C2 Server |
High/Med/Low |
Domains
| Domain |
Context |
Confidence |
|
C2 Domain |
High/Med/Low |
URLs
| URL |
Context |
Confidence |
|
Payload Download |
High/Med/Low |
Host IOCs
Registry Keys
| Key Path |
Value |
Purpose |
| HKLM...\Run |
|
Persistence |
Mutexes
| Mutex Name |
Purpose |
|
Infection marker |
File System Artifacts
| Path |
Description |
|
Dropped payload |
YARA Matches
VirusTotal Results
| Metric |
Value |
| Detection Ratio |
X / Y |
| Threat Label |
|
| First Submission |
|
| Community Score |
|
MITRE ATT&CK Mapping
| Technique |
Name |
Evidence |
| T1059.001 |
PowerShell |
Embedded PS commands |
| T1547.001 |
Registry Run Keys |
HKLM Run key modification |
| T1071.001 |
Web Protocols |
HTTP C2 communication |
Recommendations
- Block: Add network IOCs to firewall/proxy blocklists
- Detect: Deploy YARA rules on endpoints and email gateways
- Hunt: Search for host IOCs across the environment
- Share: Upload IOCs to MISP with TLP classification
references/api-reference.md (verbatim)
API Reference — Performing Malware IOC Extraction
Libraries Used
- re: Regex patterns for 16 IOC types including defanged indicators
- hashlib: MD5, SHA1, SHA256 file hashing
- pathlib: File reading (text and binary)
CLI Interface
python agent.py text --file threat_report.txt
python agent.py hash --file malware.exe
python agent.py strings --file malware.exe [--min-length 6]
python agent.py report --file malware.exe [--output iocs.json]
Core Functions
Handles defanged indicators: [.] -> ., hxxp -> http. Filters private IPs.
hash_file(file_path) — Calculate MD5/SHA1/SHA256 hashes
Extracts ASCII and wide (UTF-16LE) strings. Identifies suspicious API calls and keywords.
generate_ioc_report(file_path, output) — Full analysis report
IOC Pattern Types (16)
| Type |
Example |
| ipv4 |
192.168.1.1 (private filtered) |
| domain |
evil.example.com |
| url |
https://malware.example.com/payload |
| md5/sha1/sha256 |
File hashes |
| cve |
CVE-2024-12345 |
| registry_key |
HKLM\Software... |
| file_path_windows |
C:\Windows\Temp\mal.exe |
| mutex |
Global\MutexName |
| mitre_technique |
T1059.001 |
| bitcoin_addr |
Bitcoin wallet address |
| user_agent |
Mozilla/5.0 strings |
Suspicious String Keywords
CreateRemoteThread, VirtualAlloc, WriteProcessMemory, LoadLibrary,
GetProcAddress, WinExec, ShellExecute, powershell, cmd.exe
Dependencies
No external packages — Python standard library only.
references/standards.md (verbatim)
Standards and Frameworks Reference
IOC Types and Classification
File-Based IOCs
| Type |
Description |
Example |
| MD5 |
128-bit hash |
d41d8cd98f00b204e9800998ecf8427e |
| SHA-1 |
160-bit hash |
da39a3ee5e6b4b0d3255bfef95601890afd80709 |
| SHA-256 |
256-bit hash |
e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855 |
| Imphash |
Import hash |
PE import table hash for family grouping |
| SSDeep |
Fuzzy hash |
Context-triggered piecewise hash for similarity |
| TLSH |
Trend Micro LSH |
Locality-sensitive hash for near-duplicate detection |
Network IOCs
| Type |
Description |
Example |
| IPv4 Address |
C2 server IP |
192.0.2.1 |
| Domain |
C2 domain |
malware-c2.example.com |
| URL |
Full URL path |
https://evil.com/payload.exe |
| JA3/JA3S |
TLS fingerprint |
Client/server TLS handshake hash |
| JARM |
TLS server fingerprint |
Active TLS server scanning fingerprint |
| User-Agent |
HTTP User-Agent |
Custom UA strings in beacons |
Host-Based IOCs
| Type |
Description |
Example |
| Mutex |
Named mutex |
Global{GUID} |
| Registry Key |
Registry modification |
HKLM\SOFTWARE...\Run |
| Scheduled Task |
Persistence task |
schtasks /create ... |
| Service Name |
Malicious service |
Malicious service installation |
| Named Pipe |
IPC mechanism |
\.\pipe\name |
| PDB Path |
Debug path |
C:\Users\dev\project.pdb |
STIX 2.1 Indicator Patterns
Pattern Syntax
[file:hashes.'SHA-256' = 'abc123...']
[ipv4-addr:value = '1.2.3.4']
[domain-name:value = 'evil.com']
[url:value = 'https://evil.com/payload']
[file:name = 'malware.exe']
[email-addr:value = 'attacker@evil.com']
[network-traffic:dst_ref.type = 'ipv4-addr' AND network-traffic:dst_port = 443]
YARA Rule Structure
rule RuleName {
meta:
author = "Analyst"
description = "Detection rule"
reference = "URL"
date = "YYYY-MM-DD"
hash = "SHA256"
tlp = "white"
strings:
$text = "string" ascii wide nocase
$hex = { 4D 5A 90 00 }
$regex = /pattern[0-9]+/
condition:
uint16(0) == 0x5A4D and filesize < 5MB and any of them
}
- DOS Header: MZ signature (0x5A4D)
- PE Header: PE signature, machine type, timestamp
- Optional Header: Entry point, image base, subsystem
- Section Table: .text, .data, .rdata, .rsrc, .reloc
- Import Table: DLLs and functions used
- Export Table: Functions exported (DLLs)
- Resource Table: Embedded resources (icons, strings, configs)
References
references/workflows.md (verbatim)
Malware IOC Extraction Workflows
Workflow 1: Static Analysis Pipeline
[Malware Sample] --> [Hash Generation] --> [PE Parsing] --> [String Extraction] --> [IOC Filtering]
|
v
[YARA Scanning]
|
v
[STIX Bundle]
Steps:
- Sample Acquisition: Obtain sample from MalwareBazaar, VirusTotal, or incident response
- Hash Calculation: Generate MD5, SHA-1, SHA-256, imphash, ssdeep hashes
- PE Analysis: Parse headers, sections, imports, exports, resources, timestamps
- String Extraction: Extract ASCII/Unicode strings, apply IOC regex patterns
- IOC Filtering: Remove false positives (private IPs, common DLLs, benign domains)
- YARA Classification: Scan with community and custom YARA rules
- Output: Generate STIX 2.1 bundle with extracted indicators
Workflow 2: Dynamic Analysis Pipeline
[Malware Sample] --> [Sandbox Submission] --> [Detonation] --> [Artifact Collection]
|
+------------+------------+
| | |
v v v
[Network] [File Sys] [Registry]
[PCAPs] [Changes] [Changes]
| | |
+------------+------------+
|
v
[IOC Consolidation]
Steps:
- Sandbox Setup: Configure isolated VM with network monitoring
- Sample Submission: Submit to CAPE/Cuckoo sandbox with execution parameters
- Execution Monitoring: Monitor for 3-5 minutes of runtime behavior
- Network Capture: Extract DNS queries, HTTP/HTTPS traffic, raw connections
- File System Analysis: Identify created, modified, and deleted files
- Registry Analysis: Capture registry key changes for persistence indicators
- Process Analysis: Document spawned processes, injections, privilege escalation
- Consolidation: Merge static and dynamic IOCs into unified report
Workflow 3: Automated IOC Pipeline
[Feed/Alert] --> [Auto-Download] --> [Static Analysis] --> [Sandbox] --> [Enrichment] --> [Share]
|
v
[VirusTotal Check]
|
v
[MISP/OpenCTI Upload]
Steps:
- Trigger: New sample from malware feed, email gateway, or EDR alert
- Download: Retrieve sample securely to analysis infrastructure
- Static Scan: Automated PE parsing, string extraction, YARA scanning
- Dynamic Analysis: Submit to sandbox for behavioral analysis
- Enrichment: Check hashes against VirusTotal, cross-reference with TI platforms
- Deduplication: Remove already-known IOCs from output
- Sharing: Upload new IOCs to MISP/OpenCTI for team consumption
Back to mukul975/Anthropic-Cybersecurity-Skills (817 security skills) or Agent skills.