This server is an AI-powered study assistant for Chinese civil service exam (行测/Xingce) preparation, built on a knowledge base of 442 method cards across four major exam modules: data analysis, logic reasoning, quantity relation, and verbal comprehension.

Core tools and capabilities:

• classify_question – Automatically classify a question stem into the appropriate exam module with confidence scores.

• search_methods – Search the 442-card knowledge base by query, with optional module filtering and configurable result count.

• get_method_card – Retrieve the full content of a specific method card by ID.

• get_source_reference – Look up source files, page numbers, confidence scores, and review flags for any method.

• route_xingce_question – Determine which module a question belongs to and recommend the appropriate solving tool or track.

• solve_data_analysis – Build a structured solving draft for data analysis problems (growth rates, ratios, averages, etc.).

• solve_logic_reasoning – Build a structured solving draft for logic reasoning problems.

• compose_xingce_analysis_prompt – Generate structured analysis prompts combining routing and method scaffolds for LLM consumption.

• compose_xingce_answer_prompt – Generate conservative, safety-constrained answer prompts with strict output schemas.

• Reasoning scaffolds – Read-only method frameworks for 6 question types: graphic reasoning, definition judgement, analogy reasoning, logic analysis, quantity relation, and verbal reasoning.

Important: The server does not produce final answers or select options — it provides structured guidance and scaffolds to empower the LLM to reason step by step. It also supports multimodal input (text or image/screenshot) and batch processing when integrated with compatible AI assistants.

1. Click on "Install Server".

2. Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.

3. In the chat, type @ followed by the MCP server name and your instructions, e.g., "@花生十三-mcp 搜索“比重变化”的方法卡"

That's it! The server will respond to your query, and you can continue using it as needed.

Here is a step-by-step guide with screenshots.