Kensaku is a cross-platform desktop search and file launcher that locates documents, code, images, and other files using local indexing. It searches filenames and file contents, including PDFs, office documents, eBooks, and more than 50 programming formats. Built-in OCR recognizes text inside screenshots and images. Semantic search allows users to find documents by describing their meaning rather than remembering exact filenames. Search queries also support natural-language dates, regular expressions, and advanced filters. Incremental indexing keeps results current without repeatedly scanning entire drives. An integrated launcher provides keyboard navigation, previews, and recent-file suggestions. AI coding assistants can access the local index through a read-only MCP interface. Kensaku supports Windows, macOS, and Linux without requiring cloud accounts or telemetry.
Features
- Millisecond-scale filename searching with incremental indexing and fuzzy matching across large file collections
- Full-text searching inside PDF, DOCX, PPTX, XLSX, EPUB, email, and more than 50 programming and data formats
- Local OCR using PP-OCRv6, Indra, or Tesseract for finding text embedded in screenshots and other images
- Semantic search and natural-language queries for locating files by meaning, modification date, or remembered content
- Keyboard-driven launcher with instant previews, recent-file suggestions, advanced filtering, and grid or list layouts
- Read-only MCP integration and command-line tools for allowing AI agents to search locally indexed files without cloud processing