← 开源
WPeace-HcH

WPeGPT

An IDA plugin for binary file analysis, powered by AI models such as OpenAI and DeepSeek.

ApplicationsCodingPython
在 GitHub 打开
增长势头
+024 小时新增 Star0.0%
1.43k
Star
195
Fork
+-1
本周
3
贡献者
创建于 2023-02-28 · 更新于 2026-10-14 · 今日第 10370 名
主要开发者
WPeace-HcHXUCharlestpsnt

WPeace-HcH/WPeGPT 排名第 10,370,有 1,425 个星标。

同类别比较

WPeace-HcH/WPeGPT 排名第 10,370,此处有 8 个同类条目;milanglacier/minuet-ai.nvim 排名第 10,368。

同类别比较
名称排名指标
codota/tabnine-vscode10,348
语言:
TypeScript
星标数:
1,431
分叉数:
200
简介:
Visual Studio Code client for Tabnine. https://marketplace.visualstudio.com/items?itemName=TabNine.tabnine-vscode
thalesgroup-cert/Watcher10,362
语言:
JavaScript
星标数:
1,430
分叉数:
218
简介:
Watcher - Open Source AI-powered Cyber Threat Intelligence & Hunting Platform. Developed with Django & React JS.
Voine/ChatWaifu_Mobile10,363
语言:
C++
星标数:
1,429
分叉数:
153
简介:
移动版二次元 AI 老婆聊天器
milanglacier/minuet-ai.nvim10,368
语言:
Lua
星标数:
1,427
分叉数:
65
简介:
💃 Dance with Intelligence in Your Code. Minuet offers code completion as-you-type from popular LLMs including OpenAI, Gemini, Claude, Ollama, Llama.cpp, Codestral, and more.
jakobhoeg/nextjs-ollama-llm-ui10,372
语言:
TypeScript
星标数:
1,425
分叉数:
330
简介:
Fully-featured web interface for Ollama LLMs
theAIGuysCode/yolov4-deepsort10,375
语言:
Python
星标数:
1,424
分叉数:
744
简介:
Object tracking implemented with YOLOv4, DeepSort, and TensorFlow.
mckaywrigley/clarity-ai10,378
语言:
TypeScript
星标数:
1,423
分叉数:
272
简介:
A Perplexity clone.
anaclumos/bing-chat-for-all-browsers10,381
语言:
TypeScript
星标数:
1,422
分叉数:
102
简介:
Enable Bing ChatGPT on Chrome and Firefox
关键参数与规格
关键参数与规格
参数值
语言Python
类别Applications
子类别Coding
关键词Data analysis
星标数1,425
分叉数195
贡献者3
创建日期2023-02-28
更新日期2026-10-14
1 天星标数0
7 天星标数-1
排名10,370
上次排名10,372
排名变化2
7 天星标排名17,162
上次 7 天星标排名18,256
7 天星标排名变化1,094
总星标排名8,431
上次总星标排名8,420
总星标排名变化-11
README 摘录

NOTE] WPeChatGPT has been renamed to **WPeGPT** starting from v3. 0 with a complete architectural redesign. WPeGPT An **IDA plugin** that integrates AI (LLM) models into binary analysis workflows. WPeGPT sends decompiled pseudocode from IDA to an AI model and writes analysis results back as IDA comments. Supports multiple AI providers (OpenAI, DeepSeek, and any OpenAI-compatible API). , XOR decryption) in Python | | Vulnerability Finding | Identify potential vulnerabilities in the current function | | Exploit Generation | Attempt to generate a PoC exploit for

vulnerable functions | Automated Analysis (WPeServer + Controller) v3. Supports multiple concurrent IDA instances. Auto-detects and attempts to decrypt encrypted C2 addresses. Delete the function of analyzing encryption and decryption. Increase the function of python restore function. Added the function of finding binary vulnerabilities in functions. Increase the function of trying to automatically generate the corresponding EXP. (The upload was not tested due to the OpenAI server lag)| |2. 2* version vulnerability related functions. 1|2023-03-07|Fix the timed

out issue of OpenAI-API. (See section ***About OpenAI-API Error Reporting***)| |2. 1** to support automatic analysis of binary files. (Package *anytree* needs to be added from this version, use *requirements. Add support for other models, you can set this using the *MODEL* variable. Support for the new version of the python openai package. (Need to update your openai package)| |2. 6|2025-02-17|Add support for DeepSeek, you need to set the variable *PLUGIN_NAME* to WPeChat

-DeepSeek and fill the API KEY into variable *model_api_key*. If you want to use the R1 model, modify variable **MODEL** = *'deepseek-reasoner'*. 0** | **2026-05-27** | **Renamed to WPeGPT. Split into modular architecture ( + + ). Introduced **WPeServer** (TCP command server) for external AI-driven automation. **Three-phase analysis pipeline** (global scan → critical path → full scan). **Three analysis modes** (light / full / vuln). Intelligent string classification (10 categories) and network IoC extraction. Function suspiciousness scoring system with stdlib filtering. Structured JSON + Markdown

report generation. Configure Edit : Set your Set Set Set for Chinese (default), for English Optionally configure , , and other options 3. Install Plugin Copy and the folder to your IDA directory, then restart IDA. **NOTE**: IDA must be configured to use **Python 3**.    **Menu bar**: Edit → WPeGPT    Automated Mode (Headless Analysis) Use the wpegpt-analyzer Skill, or run Menu bar: Edit → WPeGPT → Auto-WPeGPT Reports are saved to . Contact If you encounter issues or have questions, please

open a GitHub Issue or send an email.

常见问题

记录了多少星标和分叉?

WPeace-HcH/WPeGPT 有 1,425 个星标和 195 个分叉。

仓库排名是多少?

WPeace-HcH/WPeGPT 排名第 10,370。

记录的语言和类别是什么?

WPeace-HcH/WPeGPT 有 1,425 个星标,语言为 Python,类别为 Applications。

哪个仓库排名最接近?

WPeace-HcH/WPeGPT 与 milanglacier/minuet-ai.nvim 比较,后者排名第 10,368,有 1,427 个星标。

README

English | 中文

[!NOTE] WPeChatGPT has been renamed to WPeGPT starting from v3.0 with a complete architectural redesign. Existing installations are unaffected.

WPeGPT

An IDA plugin that integrates AI (LLM) models into binary analysis workflows. WPeGPT sends decompiled pseudocode from IDA to an AI model and writes analysis results back as IDA comments. Supports multiple AI providers (OpenAI, DeepSeek, and any OpenAI-compatible API).

Inspired by Gepetto.

AI's analysis results are for reference only — otherwise we analysts would be out of work on the spot. XD

Features

Interactive Analysis (IDA Plugin)

Feature Description
Function Analysis Analyze purpose, usage environment, and behavior of a function
Variable Rename AI-suggested renaming of function variables
Python Restore Reconstruct small functions (e.g., XOR decryption) in Python
Vulnerability Finding Identify potential vulnerabilities in the current function
Exploit Generation Attempt to generate a PoC exploit for vulnerable functions

Automated Analysis (WPeServer + Controller)

v3.0 introduces a TCP-based architecture for fully automated, headless binary analysis:

  • WPeServer — An embedded TCP server inside IDA that accepts commands from an external controller. Supports multiple concurrent IDA instances.

  • Three-Phase Analysis Pipeline — Targeted segmented intelligent analysis.

  • Three Analysis Modes:

    Mode Description Time
    light Global Scan + Critical Path Function Analysis ~2-5 min
    full Global Scan + Critical Path + Full Function Analysis ~10-30 min
    vuln Critical Path Function Vulnerability Analysis ~5-20 min
  • Intelligent String Classification — Automatically categorizes strings into 10 categories: networking, keylogging, crypto, injection, persistence, antianalysis, dropper, code execution, memory/file ops, installer framework.

  • Network IoC Extraction — Extracts IPs, domains, URLs, and ports. Auto-detects and attempts to decrypt encrypted C2 addresses.

  • Function Suspiciousness Scoring — Ranks functions by keyword matching, caller/callee relationships, size, and stdlib filtering to prioritize AI analysis.

  • Shellcode Loader Detection — Pattern-based detection of shellcode execution techniques.

  • Structured Reports — Outputs both JSON and Markdown reports to _WPeAI_Results/.

Update History

Version Date Comment
1.0 2023-02-28 Based on Gepetto.
1.1 2023-03-02 1. Delete the function of analyzing encryption and decryption.
  1. Increase the function of python restore function.
  2. Modified some details.| |1.2|2023-03-03|1. Added the function of finding binary vulnerabilities in functions.
  3. Increase the function of trying to automatically generate the corresponding EXP.
  4. Modified some details.
    (The upload was not tested due to the OpenAI server lag)| |2.0|2023-03-06|1. Complete the testing of v1.2 version vulnerability related functions.
  5. Switch to the latest gpt-3.5-turbo model released by OpenAI.| |2.1|2023-03-07|Fix the timed out issue of OpenAI-API. (See section About OpenAI-API Error Reporting)| |2.3|2023-04-23|Add the Auto-WPeGPT v0.1 to support automatic analysis of binary files.
    (Package anytree needs to be added from this version, use requirements.txt or pip install anytree)| |2.4|2023-11-10|1. Changed some display details.
  6. Update Auto-WPeGPT v0.2.| |2.5|2024-08-07|1. Add support for other models, you can set this using the MODEL variable. @tpsnt
  7. Support for the new version of the python openai package. (Need to update your openai package)| |2.6|2025-02-17|Add support for DeepSeek, you need to set the variable PLUGIN_NAME to WPeChat-DeepSeek and fill the API KEY into variable model_api_key.
    (The default model is DeepSeek-V3. If you want to use the R1 model, modify variable MODEL = 'deepseek-reasoner'.)| | 3.0 | 2026-05-27 | Renamed to WPeGPT. Complete architectural redesign:
  8. Split into modular architecture (WPeGPT.py + config.py + wpe_ai_controller.py).
  9. Introduced WPeServer (TCP command server) for external AI-driven automation.
  10. Three-phase analysis pipeline (global scan → critical path → full scan).
  11. Three analysis modes (light / full / vuln).
  12. Intelligent string classification (10 categories) and network IoC extraction.
  13. Function suspiciousness scoring system with stdlib filtering.
  14. Shellcode loader detection.
  15. Structured JSON + Markdown report generation. |

Install

1. Install Dependencies

pip install -r ./requirements.txt

2. Configure

Edit WPeGPT_Config/config.py:

  • Set your API_KEY
  • Set API_BASE_URL
  • Set MODEL
  • Set ZH_CN = True for Chinese (default), False for English
  • Optionally configure ANALYSIS_MODE, MAX_WORKERS, and other options

3. Install Plugin

Copy WPeGPT.py and the WPeGPT_Config/ folder to your IDA plugins/ directory, then restart IDA.

NOTE: IDA must be configured to use Python 3.

Usage

Interactive Mode (IDA Plugin)

  • Keyboard Shortcuts:

    Shortcut Action
    Ctrl+Alt+G Function analysis
    Ctrl+Alt+R Rename function variables
    Ctrl+Alt+E Vulnerability finding
    Ctrl+Alt+W Light auto analysis
  • Right-click in the pseudocode window for context menu.

  

  • Menu bar: Edit → WPeGPT

  

Automated Mode (Headless Analysis)

Use the wpegpt-analyzer Skill, or run Menu bar: Edit → WPeGPT → Auto-WPeGPT

Reports are saved to _WPeAI_Results/.

Example

How to use:

  

Function analysis results:

  

Vulnerability finding:

  

About OpenAI-API Errors

If you experience connection issues while behind a proxy:

  • Check your urllib3 version — v1.26 has known proxy issues. Fix with:
    pip uninstall urllib3
    pip install urllib3==1.25.11
    
  • Configure FORWARD_PROXY in config.py (e.g., http://127.0.0.1:7890).
  • Or use a reverse proxy by setting API_BASE_URL.

Contact

If you encounter issues or have questions, please open a GitHub Issue or send an email.

排名相邻
同类项目