DexDec is a fast, accurate Android decompilation workbench that produces readable Java/Kotlin source from APK/DEX bytecode and integrates with AI agents via MCP.

Stars

5

7-day growth

No data

Forks

1

Open issues

0

License

Apache-2.0

Last updated

2026-07-30

AI repository intelligence
FR-AI / ANALYSIS

Why it is worth attention

It focuses on large, obfuscated APKs, provides higher-fidelity decompilation than traditional tools, and uniquely offers MCP integration for AI-assisted reverse engineering.

Who it is for

  • Android reverse engineers
  • Security researchers analyzing obfuscated apps
  • Developers who need to inspect third-party APKs
  • AI agent developers for code analysis tools

Use cases

  • Decompiling large multi-DEX applications quickly
  • Analyzing heavily obfuscated Android apps with better fidelity
  • Connecting an AI coding agent (e.g., Claude Code, Cursor) to live decompilation context
  • Searching and navigating decompiled code with symbol awareness

Strengths

  • Faster than traditional decompilers for large APKs due to on-demand source generation
  • Higher-fidelity source with fewer artifacts like broken exception blocks or artificial variables
  • Resilient – continues decompilation even when individual classes fail
  • Supports both Java and Kotlin output, switchable without reopening archive

Considerations

  • Requires Rust toolchain and Node.js to build from source
  • Desktop app currently only demonstrated for macOS (DMG build)
  • MCP integration may require compatible AI clients; not all agents supported out of the box

README quick start

Fast, accurate Android decompilation for humans and AI agents. Explore APK and DEX bytecode as readable Java or Kotlin, then hand the same live context to your coding agent.

DexDec is a native reverse-engineering workbench for large, complex, and heavily obfuscated Android applications. It opens projects quickly, produces higher-fidelity source than previous-generation Java decompilers, and keeps making progress on APKs that cause older tools to stop.

Use it directly as a focused desktop application, or connect an AI agent through MCP. The agent can understand what is open in DexDec, inspect only the code it needs, follow exact references, read packaged resources, and reveal its findings back in the desktop UI.

Features

Fast on large APKs

Start exploring large multi-DEX applications immediately. DexDec generates source only when it is needed, keeping project opening, search, and navigation responsive.

Higher-fidelity source

Recover code that is closer to what a developer would write, with fewer duplicated conditions, broken exception blocks, artificial temporary variables, and misleading types than traditional Java decompilers.

Built to keep going

One difficult class should not make an entire application unusable. DexDec isolates troublesome code and continues recovering the rest of the project instead of turning one failure into an all-or-nothing result.

Java and Kotlin

Choose the output that fits the task and switch languages without reopening the archive. Use familiar Java for compatibility or Kotlin for modern Android work.

Complete APK exploration

Browse classes, packages, AndroidManifest.xml, decoded XML, AIDL, images, resource tables, native libraries, and other packaged files in one workspace.

Symbol-aware navigation

Quick-open classes and members, jump through declarations and usages, inspect references, reveal symbols in the Explorer, and move through navigation history.

Professional workspace

Use preview and pinned tabs, split editor groups, code folding, outlines, context menus, configurable keymaps, and light or dark editor themes.

Reversible renaming

Rename classes, fields, methods, and local symbols without modifying the input. Undo or redo edits and persist project state to a DexDB file when ready.

AI-native analysis

Co

Description

Fast, accurate Android decompilation for humans and AI agents.

Related repositories

Similar projects matched by category, topics, and programming language.

0xwilliamortiz
Featured
0xwilliamortiz GitHub avatar

openclaude-improved

OpenClaude is an open-source CLI coding agent that runs on any platform and supports a wide range of LLM providers, offering the same tools and workflows as Claude Code.

AI & Machine LearningLarge Language Models
577
jamesob
Featured
jamesob GitHub avatar

local-llm

A comprehensive guide for building and configuring a high-end local machine to run state-of-the-art LLMs, with detailed hardware choices, BIOS tuning, and Docker-based model serving.

AI & Machine LearningLarge Language Models
1,660
makecindy
Featured
makecindy GitHub avatar

cindy

Cindy is an open-source AI agent that runs locally on your machine, integrates multiple AI harnesses and models, and provides memory, skills, and automation to perform real work in your projects and apps.

AI & Machine LearningLarge Language Models
958