Privesc Linux logo

Privesc Linux

Community
hypnguyen1209
privesc-linux

Use when escalating privileges on a Linux host — SUID/SGID & GTFOBins, sudo LPE (CVE-2025-32462/32463), capabilities & LD_PRELOAD, kernel LPE (CVE-2024-1086, Dirty Pipe, GameOver(lay)), service misconfig (PwnKit, Looney Tunables), container/namespace escape

Overview

Publisherhypnguyen1209
Repositoryoffensive-claude
Skill nameprivesc-linux
Stars
396
Forks
71
Bundled files
10
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 10 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by hypnguyen1209 on GitHub. Read the source before you install it.

Installation

Install the Privesc Linux AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/hypnguyen1209/offensive-claude/tree/main/skills/privesc-linux
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/hypnguyen1209/offensive-claude.git /tmp/offensive-claude
mkdir -p .claude/skills
cp -r /tmp/offensive-claude/skills/privesc-linux .claude/skills/privesc-linux
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Privesc Linux in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Privesc Linux on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Privesc Linux is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Linux Privilege Escalation

When to Activate

  • Gained an initial unprivileged shell on a Linux host and need root or a higher-privileged account
  • Post-exploitation lateral/vertical movement on Linux servers, workstations, CI runners, or appliances
  • Container / Kubernetes pod foothold that needs to escape to the host node
  • Triaging a host for misconfig-based LPE (SUID, sudo, capabilities, writable units/cron) before reaching for kernel 0-day
  • CTF / lab challenges requiring privilege escalation with a defensible, detection-aware methodology

Technique Map

TechniqueATT&CKCWEReferenceScript
Automated + manual enumeration (LinPEAS/pspy/LES2)T1082, T1057CWE-200references/enumeration-tooling.mdscripts/linpriv_enum.py
Detection-aware / low-noise enumerationT1082CWE-200references/enumeration-tooling.mdscripts/linpriv_enum.py
SUID/SGID binary abuse (GTFOBins)T1548.001CWE-269references/suid-sudo-capabilities.mdscripts/cap_suid_hunter.sh
sudo misconfig + GTFOBins escapeT1548.003CWE-269references/suid-sudo-capabilities.mdscripts/cap_suid_hunter.sh
sudo host option LPE (CVE-2025-32462)T1548.003CWE-863references/suid-sudo-capabilities.mdscripts/sudo_cve_2025_check.sh
sudo chroot/NSS LPE (CVE-2025-32463)T1548.003CWE-829references/suid-sudo-capabilities.mdscripts/sudo_cve_2025_check.sh
Linux capabilities abuse (setuid/dac_read/sys_admin)T1548CWE-250references/suid-sudo-capabilities.mdscripts/cap_suid_hunter.sh
LD_PRELOAD / LD_LIBRARY_PATH sudo hijackT1574.006CWE-426references/suid-sudo-capabilities.mdscripts/cap_suid_hunter.sh
nf_tables double-free LPE (CVE-2024-1086)T1068CWE-416references/kernel-exploits.mdscripts/kernel_exploit_suggester.py
io_uring memory-sharing LPE (CVE-2024-0582/2025-21836)T1068CWE-416references/kernel-exploits.mdscripts/kernel_exploit_suggester.py
Dirty Pipe page-cache overwrite (CVE-2022-0847)T1068CWE-787references/kernel-exploits.mdscripts/kernel_exploit_suggester.py
GameOver(lay) OverlayFS (CVE-2023-2640/32629)T1068CWE-269references/kernel-exploits.mdscripts/kernel_exploit_suggester.py
udisks/libblockdev loop-mount LPE (CVE-2025-6019)T1068CWE-250references/service-misconfig-lpe.mdscripts/linpriv_enum.py
PAM allow_active bypass (CVE-2025-6018)T1068CWE-863references/service-misconfig-lpe.mdscripts/linpriv_enum.py
glibc ld.so Looney Tunables (CVE-2023-4911)T1068CWE-787references/service-misconfig-lpe.mdscripts/kernel_exploit_suggester.py
polkit pkexec PwnKit (CVE-2021-4034)T1548.001CWE-787references/service-misconfig-lpe.mdscripts/linpriv_enum.py
Cron/systemd/PATH/writable-file abuseT1053.003, T1574.007CWE-732references/service-misconfig-lpe.mdscripts/linpriv_enum.py
NFS no_root_squash SUID dropT1222.002CWE-732references/service-misconfig-lpe.mdscripts/linpriv_enum.py
runc fd-leak container escape (CVE-2024-21626)T1611CWE-668references/container-namespace-escape.mdscripts/container_escape_check.sh
Docker socket / privileged container escapeT1611, T1610CWE-269references/container-namespace-escape.mdscripts/container_escape_check.sh
cgroup release_agent / CAP_SYS_ADMIN escapeT1611CWE-269references/container-namespace-escape.mdscripts/container_escape_check.sh
Kubernetes pod/token escapeT1611CWE-668references/container-namespace-escape.mdscripts/container_escape_check.sh

Quick Start

bash
# 0. Stabilize shell + baseline context
python3 -c 'import pty;pty.spawn("/bin/bash")'; export TERM=xterm
id; uname -a; cat /etc/os-release

# 1. Fast, detection-aware enumeration (custom, no external download needed)
python3 linpriv_enum.py --quick            # quick wins triage
python3 linpriv_enum.py --full --json out.json   # full sweep -> JSON evidence

# 2. SUID / capabilities / sudo triage (maps directly to GTFOBins)
./cap_suid_hunter.sh                        # ranks exploitable SUID + caps + sudo -l

# 3. sudo 2025 LPE check (no creds needed; default-config killers)
./sudo_cve_2025_check.sh                    # tests CVE-2025-32462 / -32463 + PwnKit

# 4. Kernel + glibc CVE mapping for current host
python3 kernel_exploit_suggester.py        # uname/glibc/distro -> ranked modern LPEs

# 5. If containerized, check escape surface
./container_escape_check.sh                 # runc fd-leak, docker.sock, caps, k8s token

# 6. Validate root, then drop SUID backup or stable persistence per ROE
id; cp /bin/bash /tmp/.b && chmod 4755 /tmp/.b   # only if authorized

OPSEC & Detection (summary)

TechniqueTelemetry / IOCDetection (Sigma/EDR)OPSEC note
LinPEAS / mass enumBurst of find / -perm, hundreds of reads, which/getcap spamauditd execve of linpeas; EDR file-scan anomalyPrefer targeted enum (linpriv_enum.py --quick); avoid full FS walks on EDR hosts
SUID/GTFOBins escapesetuid() to 0 from non-root parent, shell with -pauditd uid-change without login; Falco Run shell untrustedUse the least-noisy binary; many GTFOBins one-liners are flagged by name
sudo CVE-2025-32463syslog sudo ... CHROOT=, NSS libnss_*.so from world-writable pathElastic "Potential CVE-2025-32463 Sudo Chroot Execution"; auditd -R usagechroot dir + fake nsswitch.conf are durable IOCs — clean the tree
sudo CVE-2025-32462sudo log with -h/HOST mismatch vs real hostnameauditd sudo with --host not paired with -lleaves clean sudo log entry; blends with normal sudo
nf_tables CVE-2024-1086unshare/CLONE_NEWUSER + nftables from non-root; dmesg slab/UAFFalco Unprivileged Delegation of Page Faults; auditd unshare+nftnamespace creation is logged; disable userns to neutralize
io_uring LPEio_uring_setup syscall from unexpected procauditd syscall=io_uring_setup; eBPF LSMmany distros now ship io_uring_disabled=2
Looney Tunables CVE-2023-4911setuid exec with GLIBC_TUNABLES= containing =; core dumpsElastic "Potential Privilege Escalation via CVE-2023-4911"env var is recorded in auditd execve; unset before exec where possible
udisks CVE-2025-6019D-Bus Filesystem.Resize/Check, loop mount in /tmp w/o nosuidauditd mount w/o nosuid; Falco mount-from-loopleaves loop device + XFS image; detach + shred image
pkexec PwnKitpkexec with argc==0, GCONV_PATH= env, /var/... GConv moduleauditd pkexec + empty argv; Sigma proc_creation_lnx_pkexecdropped GConv .so + dir are IOCs; remove them
Container escape (runc)/proc/self/fd/* cwd, host paths from container, runc exec anomaliesFalco Container escape/Mount launched in container; CrowdStrike CWPescapes are heavily monitored in CWP — confirm scope before running
cron/PATH/writable unitnew file in /etc/cron*, systemctl daemon-reload, PATH-prepended binauditd watch on /etc/cron*,/etc/systemd/*; Sigma cron tamperrevert file mtimes / remove dropped scripts post-exploit

Detailed Sigma rules, auditd rule snippets and IOC lists live inside each reference file's Detection subsection.

Deep Dives

  • references/enumeration-tooling.md — LinPEAS/pspy/LES2/unix-privesc-check, the linpriv_enum.py methodology, quick-win checklists, and low-noise enumeration for EDR-monitored hosts.
  • references/suid-sudo-capabilities.md — SUID/SGID + GTFOBins, sudo misconfig escapes, sudo CVE-2025-32462 (host) and CVE-2025-32463 (chroot/NSS), Linux capabilities (cap_setuid/cap_dac_read_search/cap_sys_admin), and LD_PRELOAD/LD_LIBRARY_PATH sudo hijacking.
  • references/kernel-exploits.md — nf_tables CVE-2024-1086 (and 2026 nftables UAFs), io_uring CVE-2024-0582 / CVE-2025-21836, Dirty Pipe CVE-2022-0847, GameOver(lay) CVE-2023-2640, plus a kernel-CVE selection methodology and ROP/commit_creds primer.
  • references/service-misconfig-lpe.md — udisks/libblockdev CVE-2025-6018/6019 chain, Looney Tunables CVE-2023-4911, PwnKit CVE-2021-4034, polkit/D-Bus abuse, cron/systemd/PATH hijacking, writable /etc/passwd, and NFS no_root_squash.
  • references/container-namespace-escape.md — runc Leaky Vessels CVE-2024-21626, Docker socket / privileged-container / --pid=host / CAP_SYS_PTRACE escapes, cgroup-v1 release_agent, user namespaces, and Kubernetes service-account-token → pod escape.

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Privesc Linux AI skill do?

Use when escalating privileges on a Linux host — SUID/SGID & GTFOBins, sudo LPE (CVE-2025-32462/32463), capabilities & LD_PRELOAD, kernel LPE (CVE-2024-1086, Dirty Pipe, GameOver(lay)), service misconfig (PwnKit, Looney Tunables), container/namespace escape

Why use Privesc Linux on TypingMind?

Because you install it once and use it with any model. Privesc Linux is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Privesc Linux in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/hypnguyen1209/offensive-claude/tree/main/skills/privesc-linux. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Privesc Linux?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Privesc Linux?

As many as you like. As long as a model supports skills, you can use Privesc Linux with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Privesc Linux AI skill free?

Yes. It is published on GitHub by hypnguyen1209 under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇