Skip to content
KitploitKITPLOIT
ToolsExploitsBlog
Log in
Submit
ToolsExploitsBlog
Submit

Hacking, PenTest, and Cybersecurity Tools for Your Security Arsenal!

Kitploit is a directory of hacking, cybersecurity, and pentesting tools. Discover the latest project updates to find vulnerabilities, analyze systems, automate testing, and strengthen your security.

FeedsContactPrivacy© 2026 Kitploit

Tool Directory

Categories

View all categories
Loading categories
MaSSHole — Detection-only shell checker for the OpenSSH musl password-auth bypass (CVE-2026-123158): audits local hosts, remote hosts over SSH, or probes address ranges without credentials. | Kitploit
Tools/GitHubGitHub/liamromanis101/masshole
ReconnaissanceVulnerability ScannersConfiguration AuditingNetwork SecurityDevSecOpsAuthenticationIncident ResponseLabs & Practice
GitHubliamromanis101/masshole

MaSSHole

Detection-only shell checker for the OpenSSH musl password-auth bypass (CVE-2026-123158): audits local hosts, remote hosts over SSH, or probes address ranges without credentials.

12 days agoNot yet reviewed
Share
View Repository

MaSSHole

MaSSHole (MaSSHole.sh) is a detection-only checker for *sshole (CVE-2026-123158), the OpenSSH password-authentication bypass on musl-based Linux systems such as Alpine.

It can audit the machine it runs on, audit other machines over SSH, or probe a whole address range without logging in. It never sends a password and never attempts the bypass.

To try it against known-good and known-bad hosts first, see the Docker test lab.


The vulnerability in brief

On musl libc, OpenSSH's own password check (used when sshd runs without PAM) mishandles accounts whose /etc/shadow field is *, meaning "no password". A password between 257 and 1024 bytes is accepted for those accounts. It was fixed in OpenSSH 10.6p2, and Alpine backported the fix to its stable branches.

A host is exposed when all of these are true:

ConditionTypical on Alpine?
sshd is linked against musl libcYes
sshd runs without PAM (openssh-server package, or UsePAM no)Yes, default package
PasswordAuthentication yes for the accountYes, OpenSSH default
An account has * as its shadow password field and can log inroot:* by default, alpine:* on cloud images
At least one account has a SHA-256/512 ($5$/$6$) hashCommon
OpenSSH is older than the fixed version for that distro branchUntil you upgrade

Root is usually protected by the default PermitRootLogin prohibit-password. glibc distributions, PAM-enabled sshd, Dropbear and non-Linux systems are not affected.

Fixed versions

Alpine branchFixed openssh package
edge10.6_p2-r0
3.2410.3_p1-r2
3.2310.2_p1-r1
3.2210.0_p1-r11
3.219.9_p2-r1
3.209.7_p1-r6
3.16–3.19Patched too; check the Alpine advisory for exact versions
Upstream / other musl buildsOpenSSH 10.6p2

Requirements

  • Any POSIX sh. It works with BusyBox ash (Alpine), dash and bash, and needs only standard tools (awk, sed, grep).
  • Local mode: run as root for complete results, because it reads /etc/shadow and runs sshd -T.
  • Network and remote modes: an OpenSSH ssh client. timeout and ip are used when present.
  • Remote mode: key-based SSH access to the targets (BatchMode is used), plus root or a passwordless sudo/doas rule on them.

Install

chmod +x MaSSHole.sh
./MaSSHole.sh --help

Nothing needs installing on the targets for remote mode, because the script is piped over SSH and run there.


Modes

--local: audit this machine

sudo ./MaSSHole.sh --local

It checks, in order:

  1. Which sshd binary is running and its OpenSSH version.
  2. Whether it's linked against musl (glibc means not affected).
  3. The patch level: the installed Alpine package against the fixed version for your branch, or the upstream version for custom builds and other musl distros. It also warns if a running sshd predates an upgrade and needs a restart.
  4. The effective config from sshd -T: PAM, PasswordAuthentication, PermitRootLogin, Allow/Deny rules.
  5. The accounts with a * (or other short) shadow field. Each one is re-evaluated with sshd -T -C user=…, so Match blocks count.
  6. Indicators of compromise: Accepted password for <password-less account> in /var/log/messages*, auth.log*, secure*, logread and journalctl. Skip this with --no-logs.

--remote: audit other machines over SSH

Runs the local audit on each destination and collects the verdicts.

./MaSSHole.sh --remote -d 10.0.0.0/28 -l root
./MaSSHole.sh --remote -d 10.0.0.0/28 -l alpine --elevate doas -i ~/.ssh/fleet_key

When there is more than one host, full per-host reports are saved to masshole-reports-<timestamp>/, or to the folder given with --outdir.

--network: unauthenticated probe

Connects, reads the SSH banner, and asks the server which authentication methods it offers to each probe user (using the SSH none method). No credentials are sent.

./MaSSHole.sh --network -d 10.0.0.0/24 --csv scan.csv

From the outside, a fixed Alpine host looks identical to a vulnerable one, because Alpine backports keep the old version string. Network mode therefore reports candidates. Confirm them with --remote or --local.

Why network and remote results differ

Expect the two modes to disagree on the same hosts. That's by design, not a fault: they can see very different amounts.

Condition--network can see it?--remote / --local can see it?
OpenSSH versionPartly: banner only, often without the p1/p2 suffixYes
Distro (Debian/Ubuntu suffix)Sometimes: only distros that add a banner suffixYes
musl or glibcNoYes
PAM in useNoYes, from sshd -T
Patched Alpine package (backport)No: the version string doesn't changeYes, from the apk database
Password auth offeredYes, for the probe usersYes, per account including Match rules
* accounts and SHA-256/512 hashesNoYes, from /etc/shadow
Signs of past compromiseNoYes, from the logs

So network mode can rule hosts out reliably (glibc distros, Dropbear, password auth off, OpenSSH 10.7 and later), but it can only flag the rest as candidates. Fixed Alpine hosts, PAM builds and already-compromised hosts all look the same as a vulnerable host from outside. Remote mode separates them.

Typical results for the same host:

Host--network--remote
Unpatched Alpine, password auth onCANDIDATEVULNERABLE
Patched Alpine packageCANDIDATEOK (patched)
Alpine with PAMCANDIDATEOK (PAM in use)
Unpatched, PasswordAuthentication noNO_PASSWORD_AUTHAT_RISK (upgrade anyway)
Debian / UbuntuNOT_AFFECTEDOK (glibc)
Unpatched and already exploitedCANDIDATECOMPROMISED

Suggested workflow: sweep the range with --network (no credentials needed), then run --remote against the CANDIDATE hosts to confirm. The lab shows this with real output.


Source and destination

OptionMeaning
-s, --source ADDR|IFACELocal address (ssh -b) or interface (ssh -B, OpenSSH 8.9+) to send from. It is checked to exist before scanning.
-d, --dest TARGETSDestinations; can be repeated. Comma- or space-separated.
-f, --dest-file FILEDestinations from a file, one or more per line, with # comments.
-p, --port NDefault port. Network mode uses 22; remote mode uses your ssh config.

Destination formats can be mixed freely:

host            host:2222          alpine@host        [email protected]:2222
10.0.0.5        [2001:db8::1]:22   2001:db8::1
10.0.0.0/24     (CIDR; network and broadcast addresses skipped below /31)
10.0.0.10-50    (last-octet range)
10.0.0.10-10.0.1.20   (full range)

All options

Download Tool

Most Popular

View all →

Discover the most used tools by our community.

  • 1

    OsintHunter

    letox4/osinthunter

    🎯 The ultimate OSINT toolkit: investigation graph, confidence scoring, alias families, map, timeline and reports. 49 modules, 400+ sites, no API keys required.

    1.0k
  • 2

    ghostlock-app

    yukonga/ghostlock-app

    GhostLock One-Tap Execution App (CVE-2026-43499)

    593
  • 3

    looking-glass

    anonghosty/looking-glass

    CSIRT platform that ingests Shadowserver reports into MongoDB and replays them as interactive maps, analysis, findings, advisories and executive dashboards.

    371
  • 4

    redforge

    cybersecurity91/redforge

    Browser-side red team toolkit with 102 offline tools for recon, payload generation, credential attacks, crypto, encoding and engagement reporting.

    336
  • 5

    ARTEX

    baarai/artex

    LLM multi-agent autonomous penetration testing platform with planner/worker agents, asset and exploration graphs, MITM traffic recording, and vulnerability retesting.

    326
  • 6

    CyberStrikeAI

    ed1s0nz/cyberstrikeai

    The system of action for AI-native cybersecurity—where intent becomes governed execution, evidence becomes operational memory, and every operation improves the next.

    322

Explore all tools

Browse our collection of tools

View all tools →