XianyuAutoAgent

Report Abuse

Basic Information

Xianyu AutoAgent is an open source Python project that implements an AI customer service robot specifically for the Xianyu marketplace. It is designed to provide 7×24 automated on-platform customer support, combining context-aware dialogue, multi-expert routing and an automated bargaining system to handle buyer inquiries, price negotiations and technical questions. The system integrates an LLM backend (default model noted as 通义千问) and requires a model endpoint, API key and a web-session cookie string to interact with Xianyu. The repository targets deployment by developers or operators who want a platform-specific conversational agent; it includes configuration via a .env file, editable prompt templates, example images and a simple run entry point (python main.py). The README notes the project is for learning and exchange and may be discontinued by authors.

Links

Categorization

App Details

Features
The project documents an intelligent dialogue engine with lightweight conversation memory that feeds full session history into the LLM. It implements expert routing using prompt engineering and rule-based intent classification to dispatch requests to specialist agents (price negotiator, technical expert, default responder). A tiered price negotiation module is provided along with network search integration for technical support. Operational features include basic logging and monitoring, editable prompt templates in a prompts directory, and environment-driven configuration (API_KEY, MODEL_BASE_URL, MODEL_NAME, COOKIES_STR). The README also lists planned enhancements such as sentiment analysis, market price comparison, RAG knowledge base, DingTalk integration and a web management UI.
Use Cases
The repository helps teams automate marketplace customer service tasks by providing a ready architecture and examples for a domain-specific chatbot that can respond to users, manage conversation context and perform negotiated pricing workflows. It reduces manual workload by enabling continuous automated coverage, supports multi-expert coordination so specialized logic can be invoked for pricing or technical queries, and is customizable via prompt templates and environment settings. The project is runnable with Python 3.8+, demonstrates usage screenshots and logs, and includes instructions for obtaining cookies and configuring an LLM endpoint. It is positioned as a learning and implementation starter for developers building Xianyu-focused or similar e-commerce agents.

Please fill the required fields*