r0binak
Source-scoped identity with 8 associated PoCs and 8 linked vulnerabilities.
Exploit catalog results
Showing 8 PoCs on this page
GitHubr0binak/CVE-2026-46680
Repository PoCStars: 0Created 2026-05-21ExploitCVE-2026-466803 files
Analysis
Technical assessment
A Dockerfile that builds an image designed to exploit CVE-2026-46680 by adding a passwd entry mapping a large numeric UID (4294967296) to root, then setting the USER directive to that same large number, causing a container to run as root despite runAsNonRoot restrictions.
Backdoor review
No backdoor observed in reviewed code
The repository contains only a Dockerfile and a README. The Dockerfile demonstrates the CVE-2026-46680 vulnerability by adding a large numeric UID to /etc/passwd and setting the USER directive to that value. No backdoor, concealed executable, or operator-directed harm is present.
Classification basis and observed behavior
Classification basis
The Dockerfile is a complete, self-contained artifact that, when built and run on a vulnerable containerd instance, exercises the CVE-2026-46680 vulnerability to achieve privilege escalation. It actively constructs the conditions for the vulnerability (crafted /etc/passwd) and triggers the flawed USER parsing to run as root, which is the definition of exploit code.
Dockerfile:1-4README.md:2Requirements
- A vulnerable version of containerd (prior to 1.7.32, 2.0.9, 2.2.4, 2.3.1) must be used to run the container.
Dockerfile:1-4 - The container must be launched with a runAsNonRoot restriction (e.g., in Kubernetes) that the exploit is intended to bypass.
Dockerfile:1-4
Observed behavior
- The Dockerfile adds an /etc/passwd entry mapping UID 4294967296 to username 'x' with UID 0 (root).
Dockerfile:2 - The Dockerfile sets the USER directive to 4294967296, a number that cannot be parsed as a 32-bit integer, causing containerd to treat it as a username.
Dockerfile:3 - The container's CMD runs 'id' to demonstrate the effective UID, which would be 0 if the exploit succeeds.
Dockerfile:4
Behaviors behind the backdoor verdict
Observables
- Vulnerability Demonstration
- Payload withheldDemonstrates the CVE-2026-46680 runAsNonRoot evasion by mapping a large numeric string to root (UID 0).
Dockerfile:2-3
What the analysis did not establish
- The evidence packet contains only the Dockerfile and README.md; one additional text file is reported in metadata but its content is not provided, so its role cannot be assessed.
- The analysis is based solely on static review of the supplied text; the artifact was not executed, and its behavior in a live environment is not confirmed.
- One file (metadata only) was omitted from text review; its content is unknown.
- Binary files were not analyzed; none were identified in the artifact.
This review is limited to the supplied PoC code and context. It does not assert that the code works or is safe to execute.
GitHubr0binak/CVE-2025-14269
Repository PoCStars: 0Created 2025-12-17WriteupCVE-2025-142691 file
Analysis
Technical assessment
The artifact is a README file describing CVE-2025-14269, a credential caching vulnerability in Headlamp. It provides a vulnerability description and a textual proof-of-concept scenario, but contains no executable code, scanner, or exploit.
Backdoor review
No backdoor observed in reviewed code
The artifact contains only a README.md file describing a CVE and its proof-of-concept steps. No executable code, scripts, or deceptive payloads are present. The content is purely informational and does not exhibit backdoor or trojan behavior.
Classification basis and observed behavior
Classification basis
The artifact is a README file containing a vulnerability description and a textual proof-of-concept scenario. It does not include any executable code, scanner, or exploit. The content is purely descriptive, fitting the definition of a writeup.
README.md:1-10Requirements
- Headlamp version <= v0.38.0 installed with config.enableHelm: true
README.md:2README.md:6-7 - An authorized user must have previously accessed the Helm page to cache credentials
README.md:8
Observed behavior
- An unauthorized user can send a request to /clusters/main/helm/releases/list and receive a cached response containing sensitive information (keys, tokens, passwords)
README.md:9
Behaviors behind the backdoor verdict
Observables
- Documentation
- Payload withheldThe file is a standard vulnerability disclosure document with no executable instructions or hidden payloads.
README.md:1-10
What the analysis did not establish
- Only one file (README.md) was provided; no source code, scripts, or binaries were included.
- The analysis is based solely on the textual description; no dynamic or static analysis of the described behavior was performed.
- Only the README.md file was provided; no other repository files were included for review.
This review is limited to the supplied PoC code and context. It does not assert that the code works or is safe to execute.
GitHubr0binak/CVE-2025-23266
Repository PoCStars: 1Created 2025-09-02ExploitCVE-2025-232663 files
Analysis
Technical assessment
The artifact is a weaponized exploit for CVE-2025-23266 in the NVIDIA Container Toolkit. It builds a shared library (evil.so) that, when loaded via LD_PRELOAD inside a container, switches to the host network namespace and spawns a reverse shell to a configurable IP and port.
Backdoor review
No backdoor observed in reviewed code
The artifact is a proof-of-concept exploit for CVE-2025-23266 in NVIDIA Container Toolkit. It builds a shared library (evil.so) that, when loaded via LD_PRELOAD, switches to the host network namespace and connects back to a configurable IP/port to provide a reverse shell. The behavior is fully disclosed in the README and source code, with no concealed or materially misrepresented actions. The reverse shell target defaults to payload withheld (a common Docker bridge gateway) but is overridable via environment variables. No unrelated payloads, data exfiltration, or hidden persistence mechanisms are present.
Classification basis and observed behavior
Classification basis
The artifact contains complete, functional code that actively exploits the vulnerability to escape the container network namespace and establish a reverse shell. The README explicitly describes it as a 'Fully Weaponized NVIDIA Container Toolkit Exploit'.
README.md:2evil.go:1-62Requirements
- The attacker must be able to run a container with the NVIDIA Container Toolkit hooks enabled and set the LD_PRELOAD environment variable to point to the malicious shared library.
Dockerfile:10 - The malicious shared library (evil.so) must be present in the container's working directory.
Dockerfile:9
Observed behavior
- Builds a Go shared library (evil.so) with a constructor function that executes automatically when loaded.
Dockerfile:1-6evil.go:3-5 - The constructor function switches the process's network namespace to that of PID 1 (the host) using setns.
evil.go:19-34 - After escaping the container's network namespace, the exploit removes the LD_PRELOAD variable and deletes the evil.so file to clean up.
evil.go:39-40 - The exploit connects back to a configurable IP address and port (defaulting to payload withheld) and spawns a /bin/sh reverse shell, redirecting stdin, stdout, and stderr to the connection.
evil.go:42-59
Behaviors behind the backdoor verdict
Observables
- Reverse Shell
- Payload withheldThe exploit payload establishes a reverse shell to a configurable IP and port, which is the disclosed purpose of the PoC.
evil.go:51-59 - Network Namespace Escape
- Payload withheldThe exploit switches to the host network namespace to escape container isolation, consistent with the CVE description.
evil.go:22-33 - Default Connection Target
- Payload withheldDefault fallback IP and port if TARGET_IP/TARGET_PORT environment variables are not set. This is a common Docker bridge IP, not a concealed third-party address.
evil.go:44-49
What the analysis did not establish
- The analysis is based solely on the provided source code and metadata; the code was not executed, and its effectiveness or reliability is not verified.
- The evidence does not include any runtime output, network captures, or logs confirming successful exploitation.
- The artifact was not executed; analysis is based solely on static review of the provided source code.
This review is limited to the supplied PoC code and context. It does not assert that the code works or is safe to execute.