Linux Privilege Escalation

Fail

Audited by Socket on May 12, 2026

14 alerts found:

AnomalyMalwarex11Securityx2
AnomalyLOW
sub-skills/phase-1-system-enumeration.md

This fragment is a comprehensive host reconnaissance routine that enumerates OS/kernel, users/sessions, network topology/active connections, running processes/listening services, and environment variables (including PATH). While it shows no direct exfiltration, persistence, or payload execution within the provided code, the attacker-style breadth, staged “Phase 1” framing, and explicit PATH-hijacking hint indicate elevated suspiciousness in an untrusted/supply-chain context. Treat as recon tooling that could enable later compromise steps by a separate stage not shown here.

Confidence: 100%Severity: 60%
MalwareHIGH
sub-skills/phase-5-suid-binary-exploitation.md

This fragment is an explicit offensive playbook for SUID-based privilege escalation and post-exploitation. It enumerates SUID binaries, attempts root shell execution, steals password hash material from `/etc/shadow` and `/etc/passwd`, cracks credentials with `john`, and attempts persistence via setuid-root bash and/or root-privileged `/etc/passwd` entry manipulation. If a software supply-chain artifact ever runs or ships this code as part of install/build/runtime under the right permissions, it would represent a severe security risk.

Confidence: 60%Severity: 90%
SecurityMEDIUM
SKILL.md
MalwareHIGH
sub-skills/example-3-cron-job-script-hijacking.md

This fragment is a clear, step-by-step privilege-escalation exploit (cron job hijacking + SUID backdoor creation). There is no legitimate dependency/library behavior to evaluate; the actions shown are inherently malicious and would be treated as a critical security threat if embedded in supply-chain code.

Confidence: 60%Severity: 90%
MalwareHIGH
sub-skills/phase-4-sudo-exploitation.md

This fragment is highly indicative of malicious/exploitation intent: it contains a working root-shell payload via an `_init()`-executed LD_PRELOAD shared library and multiple sudo escape (GTFOBins-style) command examples that spawn `/bin/bash` or `/bin/sh` under elevated privileges. If present in a distributed package or executed in an environment with relevant sudo misconfigurations (especially preserved `LD_PRELOAD`), it could directly enable local root compromise. Treat as dangerous exploitation content.

Confidence: 60%Severity: 90%
MalwareHIGH
sub-skills/phase-3-kernel-exploits.md

This code fragment is an explicit kernel-exploitation procedure that fingerprints the system, selects known vulnerabilities, downloads attacker-controlled exploit source code, compiles it on the host, and executes it. The direct download→compile→execute chain without any validation or integrity controls strongly indicates malicious intent and presents an extreme security risk if distributed as part of a software supply chain.

Confidence: 60%Severity: 90%
MalwareHIGH
sub-skills/phase-7-cron-job-exploitation.md

This fragment is explicit attack code. It enumerates cron/timer execution sources, tampers with or drops an executable script containing a reverse shell, and relies on scheduled execution to trigger an outbound connection to an attacker-controlled IP/port. This indicates high likelihood of malware/persistence behavior and should be treated as dangerous and non-benign.

Confidence: 60%Severity: 90%
MalwareHIGH
sub-skills/phase-6-capabilities-exploitation.md

This is an attacker-oriented capability-exploitation snippet that performs local reconnaissance (`getcap -r /`) and then uses multiple interpreter/tool vectors to call `setuid(0)` and spawn `/bin/bash`, yielding a root shell when the host has the relevant capabilities set. If present in a dependency/package, it represents a critical security risk and strong malicious intent. No meaningful benign purpose is evident.

Confidence: 60%Severity: 90%
MalwareHIGH
sub-skills/reverse-shell-one-liners.md

This fragment is an explicit collection of reverse-shell payloads: each variant establishes an outbound TCP connection to an attacker-controlled IP/port and spawns an interactive /bin/bash shell with its standard streams redirected over the network. If present in a software dependency and executed during install/runtime, it would be a critical backdoor/backchannel capability. Confidence is high that the content itself is malicious; confidence is lower only because the snippet lacks surrounding package context showing whether/when it would be executed.

Confidence: 60%Severity: 90%
SecurityMEDIUM
sub-skills/phase-2-automated-enumeration.md
MalwareHIGH
sub-skills/phase-9-nfs-exploitation.md

This snippet is unequivocally weaponized exploit content: it deploys and runs a SUID-root shell payload over a writable NFS export, culminating in execution of `/bin/bash` from a privileged context. If present in any software supply chain or repository (beyond explicitly authorized security testing), it should be treated as highly malicious and removed, with associated artifacts and execution paths investigated.

Confidence: 60%Severity: 90%
MalwareHIGH
sub-skills/phase-8-path-hijacking.md

This code fragment is an explicit privilege-escalation exploit using PATH hijacking against a SUID binary: it places a malicious `service` executable in `/tmp`, prepends `/tmp` to `PATH`, triggers the SUID binary, and causes the attacker’s script to run `/bin/bash -p` for an elevated shell. This is strongly indicative of malicious exploitation content and would represent critical risk if found in a dependency or package script.

Confidence: 60%Severity: 90%
MalwareHIGH
sub-skills/example-1-sudo-to-root-via-find.md

This fragment is an explicit privilege-escalation exploit demonstrating root shell acquisition via a sudo misconfiguration: `sudo` permits `/usr/bin/find` as root, and the payload uses `find -exec /bin/bash` to spawn a root shell. If similar logic were present in automation or dependency-triggered scripts, it would be critically dangerous. As provided, it is not obfuscated and is clearly exploit-oriented, though it is only a command example rather than inspected package source.

Confidence: 60%Severity: 90%
MalwareHIGH
sub-skills/required-access.md

The content represents high-risk, malicious intent to facilitate unauthorized system access and remote control. It should be treated as disallowed content in code repositories and packaging processes; if encountered, remove and enforce strict gating and access controls to prevent misuse.

Confidence: 60%Severity: 90%
Audit Metadata
Analyzed At
May 12, 2026, 01:52 PM
Package URL
pkg:socket/skills-sh/Dokhacgiakhoa%2Fantigravity-ide%2Flinux-privilege-escalation%2F@4fe8cb35d8d38a3b0f457de94e3f62d90e507d69