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
CVE-2026-42779 — Proof-of-concept demonstrating a deserialization filter bypass in Apache MINA leading to remote code execution, with detailed root cause analysis, exploit PoCs, and remediation guidance. | Kitploit
Tools/GitHubGitHub/dinosn/cve-2026-42779
Vulnerability AnalysisExploitationWeb Application ExploitationPapers & ResearchLearning & Education
GitHubdinosn/cve-2026-42779

CVE-2026-42779

Proof-of-concept demonstrating a deserialization filter bypass in Apache MINA leading to remote code execution, with detailed root cause analysis, exploit PoCs, and remediation guidance.

View Repository
112356 months agoNot yet reviewed
Share

CVE-2026-42779 — Apache MINA Deserialization Filter Bypass to RCE

CVSS 3.1: 9.8 CRITICAL AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H CWE: CWE-502 Deserialization of Untrusted Data Reporter: Venkatraman Kumar, Securin Advisory: Apache Mailing List

Overview

Apache MINA versions 2.1.0 through 2.1.11 and 2.2.0 through 2.2.6 contain a deserialization filter bypass in AbstractIoBuffer.resolveClass(). The acceptMatchers allowlist — intended to restrict which Java classes can be deserialized — is completely skipped when ObjectStreamClass.forClass() returns null.

An attacker with network access to a MINA endpoint using ObjectSerializationCodecFactory can craft a protocol payload that bypasses the class filter, enabling full Remote Code Execution via standard Java deserialization gadget chains (e.g., Commons Collections).

This is an incomplete fix for CVE-2026-41635. The original patch was applied to the 2.0.x branch but was never backported to 2.1.x or 2.2.x due to a merge oversight.

Affected Versions

BranchVulnerableFixed
2.1.x2.1.0 – 2.1.112.1.12
2.2.x2.2.0 – 2.2.62.2.7

Root Cause

The vulnerability is in AbstractIoBuffer.resolveClass(), which handles class resolution during Java object deserialization.

MINA uses a custom serialization protocol with two class descriptor types:

  • Type 0 — non-Serializable classes, primitives, and arrays (standard Java class descriptor format)
  • Type 1 — Serializable classes (compact class name format)

In the vulnerable code, the acceptMatchers filter is only checked in the type-1 branch (when forClass() returns non-null). The type-0 branch calls Class.forName() directly, bypassing the filter entirely:

// AbstractIoBuffer.java — VULNERABLE (2.2.6)
protected Class<?> resolveClass(ObjectStreamClass desc) {
    Class<?> clazz = desc.forClass();

    if (clazz == null) {
        // BUG: No acceptMatchers check — filter completely bypassed
        return Class.forName(name, false, classLoader);
    } else {
        // Filter only applied here
        for (ClassNameMatcher matcher : acceptMatchers) { ... }
    }
}

The fix in 2.2.7 moves the filter check before the branch:

// AbstractIoBuffer.java — FIXED (2.2.7)
protected Class<?> resolveClass(ObjectStreamClass desc) {
    String className = desc.getName();

    // Filter applied FIRST, regardless of forClass() result
    if (!acceptMatchers.stream().anyMatch(m -> m.matches(className))) {
        throw new ClassNotFoundException("Class not in accept list " + className);
    }

    Class<?> clazz = desc.forClass();
    // ... safe resolution follows
}

Exploitation

Attack Flow

Attacker                                    Vulnerable MINA Server
   |                                              |
   |  1. Craft MINA payload with type-0           |
   |     descriptors for gadget chain classes      |
   |                                              |
   |  2. Send to endpoint using                   |
   |     ObjectSerializationCodecFactory -------->|
   |                                              |
   |          3. readClassDescriptor() reads type-0|
   |             → delegates to super (std Java)   |
   |                                              |
   |          4. resolveClass() sees forClass()==null
   |             → Class.forName() WITHOUT filter  |
   |                                              |
   |          5. Gadget chain fully deserialized   |
   |             → readObject() triggers chain     |
   |             → Runtime.exec() fires            |
   |                                              |
   |                               RCE ACHIEVED   |

Preconditions

  1. Target application uses IoBuffer.getObject() or ObjectSerializationCodecFactory
  2. Target has accept() configured (applications without a filter were already exploitable via CVE-2026-41635)
  3. A gadget chain library (Commons Collections, Spring, etc.) is on the classpath

Key Insight

The attacker controls the serialized byte stream. By using type-0 class descriptors (instead of type-1) for Serializable gadget chain classes, every class in the deserialization graph bypasses the acceptMatchers filter, regardless of the application's allowlist configuration.

Proof of Concept

Three PoCs demonstrate escalating impact:

PoCWhat it proves
FilterBypassPoC.javaFilter bypass for primitives, non-Serializable classes, arrays
CraftedBypassPoC.javaAttacker-crafted type-0 payloads bypass filter for ANY Serializable class
RcePoC.javaFull RCE via CC6 gadget chain through the filter bypass

1. Filter Bypass (MINA 2.2.6 — Vulnerable)

Classes not in the accept list are deserialized without restriction:

Filter bypass on vulnerable MINA 2.2.6

2. Crafted Payload — Arbitrary Class Loading

An attacker crafts MINA protocol payloads with type-0 descriptors to load any class past a String-only accept list:

Crafted payload bypass

3. Full RCE — Command Execution

CC6-variant gadget chain (HashSet → TiedMapEntry → LazyMap → ChainedTransformer → Runtime.exec()) achieves command execution through the filter bypass:

RCE confirmed on MINA 2.2.6

4. Filter Bypass (MINA 2.2.7 — Patched)

The same tests are blocked on the fixed version:

Filter bypass blocked on MINA 2.2.7

5. RCE Blocked (MINA 2.2.7 — Patched)

The gadget chain is rejected by the filter:

RCE blocked on MINA 2.2.7

Quick Start (Docker)

The fastest way to test — no JDK or Maven required:

# Clone this repo
git clone https://github.com/dinosn/CVE-2026-42779.git
cd CVE-2026-42779

# Build and run all PoCs
docker build -t cve-2026-42779 .
docker run --rm cve-2026-42779

# Run individual PoCs
docker run --rm cve-2026-42779 bypass     # Filter bypass only
docker run --rm cve-2026-42779 crafted    # Crafted payload bypass
docker run --rm cve-2026-42779 rce        # Full RCE

# Drop into a shell to explore
docker run --rm -it cve-2026-42779 shell

The image bundles the vulnerable MINA 2.2.6 JAR, Commons Collections 3.2.2, and all three pre-compiled PoCs. Everything runs self-contained inside the container.

Reproduction (from source)

If you prefer to build from source:

# Clone and build the vulnerable version
git clone https://github.com/apache/mina.git /tmp/apache-mina
cd /tmp/apache-mina
git checkout 2.2.6
mvn install -pl mina-core -DskipTests -q

# Download commons-collections (for RCE PoC)
curl -sL "https://repo1.maven.org/maven2/commons-collections/commons-collections/3.2.2/commons-collections-3.2.2.jar" \
  -o commons-collections-3.2.2.jar
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.

    911
  • 2

    ghostlock-app

    yukonga/ghostlock-app

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

    548
  • 3

    redforge

    cybersecurity91/redforge

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

    366
  • 4

    ARTEX

    baarai/artex

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

    362
  • 5

    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.

    311
  • 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.

    287

Explore all tools

Browse our collection of tools

View all tools →