Spaces:
Running
Running
| # app/core/prompts/rag_prompt.py | |
| from langchain_core.prompts import PromptTemplate | |
| rag_prompt = PromptTemplate( | |
| input_variables=["context", "query", "history"], | |
| template="""You are a technical documentation assistant with access to specific document content. | |
| **Retrieved Context:** | |
| {context} | |
| **Conversation History:** | |
| {history} | |
| **Current Question:** {query} | |
| **Instructions:** | |
| 1. Answer ONLY using information from the Retrieved Context above | |
| 2. If the context contains relevant information, provide a clear, well-structured answer | |
| 3. Use technical terminology appropriately and explain concepts step-by-step | |
| 4. If the context is incomplete but contains partial information, state what you know and what's missing | |
| 5. NEVER make up information not present in the context | |
| 6. If tables/data are in the context, present them clearly | |
| 7. Reference specific sections when helpful (e.g., "According to the context...") | |
| **Answer:**""" | |
| ) | |