mumu-0922/GankAIGC

What it solves

GankAIGC is designed to help researchers and students reduce the "AI-generated" signature in their academic papers. It provides tools for academic polishing, originality enhancement, and rewriting to help text pass AI detection systems (specifically mentioning Zhuque AI) while maintaining academic quality.

How it works

The system uses a FastAPI backend and a React frontend, connecting to OpenAI-compatible LLM APIs to rewrite and polish text. It features a durable task queue managed via PostgreSQL to handle concurrent users and avoid API rate limits. For AI detection, it integrates with the Zhuque AI platform, either through a local browser (for desktop users) or a browser-agent plugin (for VPS/Docker deployments).

Who it’s for

Academic writers, students, and researchers who need to refine their papers, enhance original expression, and ensure their work is not flagged as AI-generated by detection tools.

Highlights

• Multiple Processing Modes: Supports academic polishing, originality enhancement, and emotion-based rewriting.
• BYOK Support: Users can provide their own OpenAI-compatible API keys for processing tasks.
• Integrated AI Detection: Direct integration with Zhuque AI to check detection scores before and after rewriting.
• Project Management: Allows users to archive tasks by paper project, track history, and export results to Word (.docx) or Markdown (.md).
• Flexible Deployment: Available as source code, Docker Compose, or a one-click Windows portable package with an embedded PostgreSQL database.