
analyzing-cobalt-strike-beacon-configuration
Extract and analyze Cobalt Strike beacon configuration from PE files and memory dumps to identify C2 infrastructure, malleable profiles, and operator tradecraft.
- Installs
- 404
- Last 24h
- —
- Trending rank
- —
- Security
- Not audited
Install
Works with every agent the skills CLI supports.
npx skills add https://github.com/mukul975/anthropic-cybersecurity-skills --skill analyzing-cobalt-strike-beacon-configuration- 01
Install
Run the command in your project. The CLI asks which agents to install for.
- 02
Describe the task
Ask your agent for something this skill covers. It loads analyzing-cobalt-strike-beacon-configuration automatically when relevant.
- 03
Review the result
The agent follows the SKILL.md steps, scripts and references bundled with the skill.
SKILL.md
The instructions your agent reads when the skill is triggered.
Analyzing Cobalt Strike Beacon Configuration
Overview
Cobalt Strike is a commercial adversary simulation tool widely abused by threat actors for post-exploitation operations. Beacon payloads contain embedded configuration data that reveals C2 server addresses, communication protocols, sleep intervals, jitter values, malleable C2 profile settings, watermark identifiers, and encryption keys. Extracting this configuration from PE files, shellcode, or memory dumps is critical for incident responders to map attacker infrastructure and attribute campaigns. The beacon configuration is XOR-encoded using a single byte (0x69 for version 3, 0x2e for version 4) and stored in a Type-Length-Value (TLV) format within the .data section.
When to Use
- When investigating security incidents that require analyzing cobalt strike beacon configuration
- When building detection rules or threat hunting queries for this domain
- When SOC analysts need structured procedures for this analysis type
- When validating security monitoring coverage for related attack techniques
Prerequisites
- Python 3.9+ with
dissect.cobaltstrike,pefile,yara-python - SentinelOne CobaltStrikeParser (
parse_beacon_config.py) - Hex editor (010 Editor, HxD) for manual inspection
- Understanding of PE file format and XOR encoding
- Memory dump acquisition tools (Volatility3, WinDbg)
- Network analysis tools (Wireshark) for C2 traffic correlation
Key Concepts
Beacon Configuration Structure
Cobalt Strike beacons store their configuration as a blob of TLV (Type-Length-Value) entries within the .data section of the PE. Stageless beacons XOR the entire beacon code with a 4-byte key. The configuration blob itself uses a single-byte XOR key. Each TLV entry contains a 2-byte type identifier (e.g., 0x0001 for BeaconType, 0x0008 for C2Server), a 2-byte length, and variable-length data.
Malleable C2 Profiles
The beacon configuration encodes the malleable C2 profile that dictates HTTP request/response transformations, including URI paths, headers, metadata encoding (Base64, NetBIOS), and data transforms. Analyzing these settings reveals how the beacon disguises its traffic to blend with legitimate web traffic.
Watermark and License Identification
Each Cobalt Strike license embeds a unique watermark (4-byte integer) into generated beacons. Extracting the watermark can link multiple beacons to the same operator or cracked license. Known watermark databases maintained by threat intelligence providers map watermarks to specific threat actors or leaked license keys.
Workflow
Step 1: Extract Configuration with CobaltStrikeParser
#!/usr/bin/env python3
"""Extract Cobalt Strike beacon config from PE or memory dump."""
import sys
import json
# Using SentinelOne's CobaltStrikeParser
# pip install dissect.cobaltstrike
from dissect.cobaltstrike.beacon import BeaconConfig
def extract_beacon_config(filepath):
"""Parse beacon configuration from file."""
configs = list(BeaconConfig.from_path(filepath))
if not configs:
print(f"[-] No beacon configuration found in {filepath}")
return None
for i, config in enumerate(configs):
print(f"\n[+] Beacon Configuration #{i+1}")
print(f"{'='*60}")
settings = config.as_dict()
# Critical fields for incident response
critical_fields = [
"SETTING_C2_REQUEST",
"SETTING_C2_RECOVER",
"SETTING_PUBKEY",
"SETTING_DOMAINS",
"SETTING_BEACONTYPE",
"SETTING_PORT",
"SETTING_SLEEPTIME",
"SETTING_JITTER",
"SETTING_MAXGET",
"SETTING_SPAWNTO_X86",
"SETTING_SPAWNTO_X64",
"SETTING_PIPENAME",
"SETTING_WATERMARK",
"SETTING_C2_VERB_GET",
"SETTING_C2_VERB_POST",
"SETTING_USERAGENT",
"SETTING_PROTOCOL",
]
for field in critical_fields:
value = settings.get(field, "N/A")
print(f" {field}: {value}")
return settings
return None
def extract_c2_indicators(config):
"""Extract actionable C2 indicators from beacon config."""
indicators = {
"c2_domains": [],
"c2_ips": [],
"c2_urls": [],
"user_agent": "",
"named_pipes": [],
"spawn_processes": [],
"watermark": "",
}
if not config:
return indicators
# Extract C2 domains
domains = config.get("SETTING_DOMAINS", "")
if domains:
for domain in str(domains).split(","):
domain = domain.strip().rstrip("/")
if domain:
indicators["c2_domains"].append(domain)
# Extract user agent
indicators["user_agent"] = str(config.get("SETTING_USERAGENT", ""))
# Extract named pipes
pipe = config.get("SETTING_PIPENAME", "")
if pipe:
indicators["named_pipes"].append(str(pipe))
# Extract spawn-to processes
for arch in ["SETTING_SPAWNTO_X86", "SETTING_SPAWNTO_X64"]:
proc = config.get(arch, "")
if proc:
indicators["spawn_processes"].append(str(proc))
# Extract watermark
indicators["watermark"] = str(config.get("SETTING_WATERMARK", ""))
return indicators
if __name__ == "__main__":
if len(sys.argv) < 2:
print(f"Usage: {sys.argv[0]} <beacon_file_or_dump>")
sys.exit(1)
config = extract_beacon_config(sys.argv[1])
if config:
indicators = extract_c2_indicators(config)
print(f"\n[+] Extracted C2 Indicators:")
print(json.dumps(indicators, indent=2))
Step 2: Manual XOR Decryption of Beacon Config
import struct
def find_and_decrypt_config(data):
"""Manually locate and decrypt beacon configuration."""
# Cobalt Strike 4.x uses 0x2e as XOR key
xor_keys = [0x2e, 0x69] # v4, v3
for xor_key in xor_keys:
# Search for the config magic bytes after XOR
# Config starts with 0x0001 (BeaconType) XOR'd with key
magic = bytes([0x00 ^ xor_key, 0x01 ^ xor_key,
0x00 ^ xor_key, 0x02 ^ xor_key])
offset = data.find(magic)
if offset == -1:
continue
print(f"[+] Found config at offset 0x{offset:x} (XOR key: 0x{xor_key:02x})")
# Decrypt the config blob (typically 4096 bytes)
config_size = 4096
encrypted = data[offset:offset + config_size]
decrypted = bytes([b ^ xor_key for b in encrypted])
# Parse TLV entries
entries = parse_tlv(decrypted)
return entries
return None
def parse_tlv(data):
"""Parse Type-Length-Value configuration entries."""
entries = {}
offset = 0
# TLV field type mapping
field_names = {
0x0001: "BeaconType",
0x0002: "Port",
0x0003: "SleepTime",
0x0004: "MaxGetSize",
0x0005: "Jitter",
0x0006: "MaxDNS",
0x0007: "Deprecated_PublicKey",
0x0008: "C2Server",
0x0009: "UserAgent",
0x000a: "PostURI",
0x000b: "Malleable_C2_Instructions",
0x000c: "Deprecated_HttpGet_Metadata",
0x000d: "SpawnTo_x86",
0x000e: "SpawnTo_x64",
0x000f: "CryptoScheme",
0x001a: "Watermark",
0x001d: "C2_HostHeader",
0x0024: "PipeName",
0x0025: "Year",
0x0026: "Month",
0x0027: "Day",
0x0036: "ProxyHostname",
}
while offset + 6 <= len(data):
entry_type = struct.unpack(">H", data[offset:offset+2])[0]
entry_len_type = struct.unpack(">H", data[offset+2:offset+4])[0]
entry_len = struct.unpack(">H", data[offset+4:offset+6])[0]
if entry_type == 0:
break
value_start = offset + 6
value_end = value_start + entry_len
value_data = data[value_start:value_end]
field_name = field_names.get(entry_type, f"Unknown_0x{entry_type:04x}")
if entry_len_type == 1: # Short
value = struct.unpack(">H", value_data[:2])[0]
elif entry_len_type == 2: # Int
value = struct.unpack(">I", value_data[:4])[0]
elif entry_len_type == 3: # String/Blob
value = value_data.rstrip(b'\x00').decode('utf-8', errors='replace')
else:
value = value_data.hex()
entries[field_name] = value
print(f" {field_name}: {value}")
offset = value_end
return entries
Step 3: YARA Rule for Beacon Detection
import yara
cobalt_strike_rule = """
rule CobaltStrike_Beacon_Config {
meta:
description = "Detects Cobalt Strike beacon configuration"
author = "Malware Analysis Team"
date = "2025-01-01"
strings:
// XOR'd config marker for CS 4.x (key 0x2e)
$config_v4 = { 2e 2f 2e 2c }
// XOR'd config marker for CS 3.x (key 0x69)
$config_v3 = { 69 68 69 6b }
// Common beacon strings
$str_pipe = "\\\\.\\pipe\\" ascii wide
$str_beacon = "beacon" ascii nocase
$str_sleeptime = "sleeptime" ascii nocase
// Reflective loader pattern
$reflective = { 4D 5A 41 52 55 48 89 E5 }
condition:
($config_v4 or $config_v3) or
(2 of ($str_*) and $reflective)
}
"""
def scan_for_beacons(filepath):
"""Scan file with YARA rules for Cobalt Strike beacons."""
rules = yara.compile(source=cobalt_strike_rule)
matches = rules.match(filepath)
for match in matches:
print(f"[+] YARA Match: {match.rule}")
for string_match in match.strings:
offset = string_match.instances[0].offset
print(f" String: {string_match.identifier} at offset 0x{offset:x}")
return matches
Step 4: Network Traffic Correlation
from dissect.cobaltstrike.c2 import HttpC2Config
def analyze_c2_profile(beacon_config):
"""Analyze malleable C2 profile from beacon configuration."""
print("\n[+] Malleable C2 Profile Analysis")
print("=" * 60)
# HTTP GET configuration
get_verb = beacon_config.get("SETTING_C2_VERB_GET", "GET")
get_uri = beacon_config.get("SETTING_C2_REQUEST", "")
print(f"\n HTTP GET Request:")
print(f" Verb: {get_verb}")
print(f" URI: {get_uri}")
# HTTP POST configuration
post_verb = beacon_config.get("SETTING_C2_VERB_POST", "POST")
post_uri = beacon_config.get("SETTING_C2_POSTREQ", "")
print(f"\n HTTP POST Request:")
print(f" Verb: {post_verb}")
print(f" URI: {post_uri}")
# User Agent
ua = beacon_config.get("SETTING_USERAGENT", "")
print(f"\n User-Agent: {ua}")
# Host header
host = beacon_config.get("SETTING_C2_HOSTHEADER", "")
print(f" Host Header: {host}")
# Sleep and jitter for traffic pattern
sleep_ms = beacon_config.get("SETTING_SLEEPTIME", 60000)
jitter = beacon_config.get("SETTING_JITTER", 0)
print(f"\n Sleep Time: {sleep_ms}ms")
print(f" Jitter: {jitter}%")
# Generate Suricata/Snort signatures
print(f"\n[+] Suggested Network Signatures:")
if ua:
print(f' alert http any any -> any any (msg:"CS Beacon UA"; '
f'content:"{ua}"; http_user_agent; sid:1000001; rev:1;)')
if get_uri:
print(f' alert http any any -> any any (msg:"CS Beacon URI"; '
f'content:"{get_uri}"; http_uri; sid:1000002; rev:1;)')
Validation Criteria
- Beacon configuration successfully extracted from PE file or memory dump
- C2 server domains/IPs correctly identified with port and protocol
- Malleable C2 profile parameters decoded showing HTTP transforms
- Watermark value extracted for attribution correlation
- Sleep time and jitter values match observed network beacon intervals
- YARA rules detect beacon in both packed and unpacked samples
- Network signatures generated from extracted C2 profile
References
FAQ
What does the analyzing-cobalt-strike-beacon-configuration skill do?
Extract and analyze Cobalt Strike beacon configuration from PE files and memory dumps to identify C2 infrastructure, malleable profiles, and operator tradecraft.
How do I install analyzing-cobalt-strike-beacon-configuration?
Run npx skills add https://github.com/mukul975/anthropic-cybersecurity-skills --skill analyzing-cobalt-strike-beacon-configuration in your project. The skills CLI detects the agents you use and copies the skill into the right folder. For Claude Code you can also place the skill folder at ~/.claude/skills/analyzing-cobalt-strike-beacon-configuration/ (personal) or .claude/skills/analyzing-cobalt-strike-beacon-configuration/ (project).
Which AI agents can use analyzing-cobalt-strike-beacon-configuration?
Any agent that supports the open Agent Skills format (a folder with a SKILL.md file), including Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI, Windsurf, Cline, OpenCode and more.
Is analyzing-cobalt-strike-beacon-configuration safe to install?
No partner security audit is available yet. Skills can run commands and scripts, so read the SKILL.md and any bundled files before installing.
Who maintains analyzing-cobalt-strike-beacon-configuration?
It is published by mukul975 in mukul975/anthropic-cybersecurity-skills. It has 404 recorded installs.
More from mukul975/anthropic-cybersecurity-skills
testing-api-security-with-owasp-top-10
mukul975/anthropic-cybersecurity-skills
Systematically assesses REST, GraphQL, and gRPC API endpoints against the OWASP API Security Top 10 (2023) using Burp Suite and Postman for automated and manual testing. Use during authorized API penetration tests, before deploying new endpoints to production, or when validating API gateway controls and rate limiting.
analyzing-api-gateway-access-logs
mukul975/anthropic-cybersecurity-skills
Parses API Gateway access logs (AWS API Gateway, Kong, Nginx) to detect BOLA/IDOR attacks, rate limit bypass, credential scanning, and injection attempts. Uses pandas for statistical analysis of request patterns and anomaly detection. Use when investigating API abuse or building API-specific threat detection rules.
analyzing-android-malware-with-apktool
mukul975/anthropic-cybersecurity-skills
Perform static analysis of Android APK malware using apktool for resource decompilation, jadx for Java source recovery, and androguard for manifest inspection, dangerous permission-combination detection, and identification of obfuscated code, dynamic code loading, and reflection-based API calls. Use to statically triage a suspicious APK without executing it or to build mobile malware detection rules.
acquiring-disk-image-with-dd-and-dcfldd
mukul975/anthropic-cybersecurity-skills
Create forensically sound bit-for-bit disk images with dd or dcfldd on a Linux forensic workstation, preserving evidence integrity through hash verification (MD5/SHA) during acquisition. Use when imaging a suspect drive, USB device, or memory card for investigation, preserving volatile disk evidence during incident response, or producing a verified copy for legal or law-enforcement proceedings before any destructive analysis.
performing-web-application-penetration-test
mukul975/anthropic-cybersecurity-skills
Performs systematic security testing of web applications following the OWASP Web Security Testing Guide (WSTG) methodology to identify vulnerabilities in authentication, authorization, input validation, session management, and business logic. The tester uses Burp Suite as the primary interception proxy alongside manual testing techniques to find flaws that automated scanners miss. Activates for requests involving web app pentest, OWASP testing, application security assessment, or web vulnerability testing.
testing-for-xss-vulnerabilities
mukul975/anthropic-cybersecurity-skills
Tests web applications for reflected, stored, and DOM-based Cross-Site Scripting by injecting JavaScript payloads with Burp Suite (XSS extensions, Active Scan++) and browser tools, then bypassing sanitization and CSP to demonstrate session hijacking and user impersonation. Use for OWASP WSTG client-side injection testing or when evaluating input sanitization and output encoding coverage.
Popular ai & agents skills
google-agents-cli-publish
google/agents-cli
This skill should be used when the user wants to "publish an agent", "publish my ADK agent", "register an agent with Gemini", "publish to Gemini" (formerly Gemini Enterprise), or needs guidance on the agents-cli publish gemini command. Also use when the user wants to "manage agents in Agent Registry", "list/update/delete registered agents", or "register an MCP server". Covers ADK vs A2A registration modes, programmatic and interactive usage, flag reference, auto-detection from deployment metadata, Agent Registry fleet management, and troubleshooting. Part of the agents-cli skills suite. Do NOT use for deployment (use google-agents-cli-deploy).
google-agents-cli-eval
google/agents-cli
This skill should be used when the user wants to "run an evaluation", "evaluate my agent", "evaluate my ADK agent", "write an eval dataset", "analyze eval failures", "compare eval results", "optimize agent", or needs guidance on the Gemini platform eval methodology and the Quality Flywheel. Covers eval metrics, dataset schema, LLM-as-judge scoring, and common failure causes. Applies to any agents-cli project, whatever framework the agent is written in. Do NOT use for agent API code patterns (ADK: use google-agents-cli-adk-code), deployment (use google-agents-cli-deploy), or project scaffolding (use google-agents-cli-scaffold).
skill-creator
anthropics/skills
Create new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.
codex-pet
prime-skills/runcomfy-agent-skills
Codex Pet generator on RunComfy. Build a Codex-compatible Codex Pet spritesheet.webp + pet.json from a single reference image, drop it into ${CODEXHOME:-$HOME/.codex}/pets/ / and Codex picks it up as a custom Codex Pet next to the 8 built-ins. This skill produces the exact Codex Pet atlas Codex expects (1536x1872 PNG/WebP, 8 cols x 9 rows, 192x208 cells, 9 animation states — idle, running-right, running-left, waving, jumping, failed, waiting, running, review). Calls OpenAI GPT Image 2 edit ONCE via the local RunComfy CLI as runcomfy run openai/gpt-image-2/edit to produce a canonical Codex Pet pose, then assembles all 9 animation rows programmatically with ImageMagick micro-transforms — no Codex Pro, no $imagegen, no OPENAIAPIKEY required, only RUNCOMFYTOKEN. Triggers on "codex pet", "create codex pet", "make codex pet", "hatch codex pet", "/hatch image", "desktop pet codex", "codex pets", "spritesheet.webp", or any explicit ask to build a custom pet for OpenAI Codex.
full-output-enforcement
leonxlnx/taste-skill
Overrides default LLM truncation behavior. Enforces complete code generation, bans placeholder patterns, and handles token-limit splits cleanly. Apply to any task requiring exhaustive, unabridged output.
executing-plans
obra/superpowers
Use when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available