File Search Server

import streamlit as st from fastmcp_file_search import search_files from models import SearchRequest st.title("🔍 Local File Search Agent") st.markdown("Search for files in your local folders by name and content") # Input fields col1, col2 = st.columns([2, 1]) with col1: folder_path = st.text_input( "Folder Path", value="/Users/lucy/Documents", help="Enter the full path to the folder you want to search" ) with col2: max_results = st.number_input( "Max Results", min_value=1, max_value=50, value=10, help="Maximum number of results to return" ) search_prompt = st.text_area( "Search Prompt", placeholder="e.g., python files with machine learning, or .pdf documents about AI", help="Describe what files you're looking for. Include file types (e.g., .py, .txt) and keywords for content search" ) # Search button if st.button("🔍 Search Files", type="primary"): if search_prompt and folder_path: request = SearchRequest( folder_path=folder_path, search_prompt=search_prompt, max_results=max_results ) with st.spinner("Searching files..."): results = search_files(request) if results and not results[0].get("error"): st.success(f"Found {len(results)} files") for result in results: with st.expander(f"📄 {result['file_path']} (Score: {result['relevance_score']})"): st.write(f"**Path:** {result['relative_path']}") #st.write(f"**Full Path:** {result['file_path']}") st.write(f"**Matches:** {result['match_details']}") else: if results and results[0].get("error"): st.error(results[0]["error"]) else: st.warning("No files found matching your search criteria") else: st.warning("Please enter both folder path and search prompt") # Help section with st.expander("📖 How to Use"): st.markdown(""" 1. **Folder Path**: Enter the full path to the folder you want to search 2. **Search Prompt**: Describe what you're looking for: - File types: `.py`, `.txt`, `.pdf`, etc. - Keywords: Content you want to find in files 3. **Max Results**: Limit the number of results returned **Examples:** - "python files with machine learning" - ".pdf documents about AI" - "txt files containing password" """) # Footer st.markdown("---") st.markdown("*Built with FastMCP and Streamlit*")