National Cyber Warfare Foundation (NCWF)login

khaos-c2 for blending agent traffic into trusted cloud services


0 user ratings
2026-11-09 13:26:59
milo
Red Team (CNA)
"khaos-c2

khaos-c2 is a modern command-and-control framework that routes agent traffic through cloud services most networks already trust, built for authorized red team engagements and adversary emulation.








Tool28Zaaky/khaos-c2 — a modern C2 framework that routes agent traffic through cloud services already trusted by most networks
CategoryCommand-and-control framework, written in C
Primary UseOperating agent command channels during authorized red team engagements and adversary emulation exercises, blending traffic with legitimate cloud providers
Safe UseFor use exclusively inside authorized penetration tests, isolated lab environments, and contracted adversary emulation engagements with written permission from system owners
Telemetry NoteAs a cloud-masking C2, its traffic appears as ordinary connections to popular cloud providers; defenders should watch for anomalous volumes or unusual request patterns toward trusted SaaS domains from endpoints with no business need

khaos-c2 positions itself in a crowded field of post-Cobalt Strike tooling by making one architectural bet explicit in its own description: agent traffic should not originate from rented VPS boxes with suspicious reputations, but from cloud services that most corporate networks already trust implicitly. The repository sits at 28Zaaky/khaos-c2, is written primarily in C, carries an MIT license, and has accumulated a modest 232 stars, which is a reasonable signal of niche interest rather than hype-driven adoption. This article is a documentary analysis of the project's publicly stated design for authorized professionals and defenders; it is not a deployment guide.


The topic list on the repository is unusually candid about lineage and intent. Alongside the expected c2, command-and-control, post-exploitation, and redteam tags, the author explicitly references cobalt-strike and havoc, effectively declaring which ecosystems khaos-c2 wants to be measured against. The presence of evasion as a first-class topic tells you the framework's value proposition is not raw capability but blendability — the ability to make agent check-ins look like ordinary cloud usage. Tags for both windows and linux suggest cross-platform implant support, or at least a server architecture that does not assume a single victim operating system.


The C implementation language is a meaningful design signal. While modern frameworks like Havoc lean on C++ with a Qt-based teamserver UI, a leaner C codebase typically implies a smaller compilation footprint, fewer runtime dependencies, and implant components that are easier to strip down for size-conscious operations. For defenders, that same property matters: a small C agent produces a small binary surface, which reduces the utility of naive signature-based detection and shifts the burden toward behavioral and network-layer telemetry.


The core architectural claim — routing agent traffic through cloud services already trusted by most networks — describes a domain-fronting-adjacent or SaaS-abuse pattern that has become the dominant trend in commercial and open-source C2 alike. Rather than beaconing directly to an operator-controlled IP, the implant communicates with infrastructure hosted behind or via major cloud providers, so the network egress looks like a session with a legitimate, allowlisted destination. This is precisely the technique category that pushed vendors toward TLS inspection, JA3/JA4 fingerprinting, and egress policy enforcement, and it is the reason tools like this are valuable study subjects for blue teams.


It is worth being transparent about the evidentiary limits here: the repository's README is empty at the time of this analysis, so every capability statement above derives from the project's own short description, its topic taxonomy, and repository metadata. That thinness is itself worth a caution flag for operators considering adoption. A C2 framework without documentation means no documented build process, no stated operational security guarantees, no threat model, and no guidance on safe handling of listeners or agent artifacts. Anyone evaluating khaos-c2 for an authorized engagement should audit the C source on the main branch before it touches engagement infrastructure, treating the code as untrusted until reviewed.


An empty README also means you cannot verify claims that often accompany new C2 projects: encryption of agent traffic, malleable profile support, sleep obfuscation, or operator authentication on the teamserver. None of those features should be assumed present. What the topics do confirm is the author's framing of the tool as post-exploitation infrastructure — meaning it presumes an initial foothold already exists and concerns itself with maintaining durable, low-visibility command channels across windows and linux targets. That scoping matters for defenders prioritizing detections: the relevant telemetry is long-lived, low-and-slow outbound connections to cloud endpoints, not initial-access indicators.


From a red team perspective, the draw of khaos-c2 is consolidation: a free, MIT-licensed, C-native framework in the lineage of Cobalt Strike and Havoc but with cloud-native traffic routing as a design primitive rather than a bolted-on profile. Commercial alternatives price domain fronting and redirector tooling into expensive licenses, so open implementations of the same idea tend to attract attention quickly — the 232-star count on a documentation-free project suggests operators are reading the source directly rather than waiting for docs.


For defensive analysts, the existence and traction of this project is a useful planning input. It reinforces that cloud-SaaS destinations should not be blanket-trusted in egress policy, and that detection engineering effort is better spent on volumetric and temporal anomalies — periodic beacons to a single cloud endpoint, uniform request sizes, unusual user agents against known-good baselines — than on IP reputation. Purple team exercises can use the published architecture pattern, in a lab, to validate whether existing SSL/TLS inspection and endpoint telemetry would surface an agent that blends into legitimate cloud traffic.


Repository hygiene is a final consideration. MIT licensing means the code can be audited, forked, and studied freely, which is good for research and bad for attribution — forks of a C2 framework can diverge quickly, and detections tuned to the upstream codebase may miss modified variants. The single main branch and low documentation level suggest an early-stage, single-maintainer project (28Zaaky), so professional adopters should expect rapid API churn and plan to pin a reviewed commit rather than tracking HEAD.


In sum, khaos-c2 is best understood as an early but conceptually current entry in the cloud-masking C2 category: C-native, cross-platform in ambition, and explicitly aligned with the redteam tooling lineage of Cobalt Strike and Havoc. For authorized professionals, its value today is as a source-auditable reference implementation of trusted-cloud traffic routing; for defenders, it is a concrete reminder that the boundary between legitimate SaaS traffic and covert channel has never been thinner, and that detection programs anchored to reputation lists are already behind the curve. Clone it into an isolated lab, read the code before you trust it, and treat the empty README as the first finding of your assessment.



Official project repository for 28Zaaky/khaos-c2.

Download Tool

Educational analysis for authorized security professionals. Use only in controlled, authorized environments.






Source: OffensiveSec
Source Link: https://www.offsecblog.com/2026/11/khaos-c2-for-blending-agent-traffic.html


Comments
new comment
Nobody has commented yet. Will you be the first?
 
Forum
Red Team (CNA)



Copyright 2012 through 2026 - National Cyber Warfare Foundation - All rights reserved worldwide.