orchata
от ehudsn
Search & ResearchКратко
Knowledge management and RAG platform with tree-based document indexing.
Описание
Что делает этот навык
Orchata is a RAG platform that organizes documents into spaces with tree-based indexing. It supports PDF, Word, Excel, PowerPoint, Markdown, and image files. AI assistants interact with knowledge bases through MCP tools for search, upload, and document management.
Tree-based document indexing lets you navigate large files by section and summary rather than retrieving unstructured chunks.
Когда его использовать
- Searching internal documentation with natural language queries
- Uploading and indexing new PDFs for team knowledge bases
- Navigating large documents by section without reading everything
- Discovering which knowledge spaces contain relevant content
- Checking processing status of recently uploaded documents
Пример сценария
User asks: Find the authentication steps in our API docs
- 1
Call smart_query with query='API authentication steps' to identify relevant spaces
- 2
Call query_spaces with the returned space IDs and query='authentication steps'
- 3
Review returned results with full content and page ranges
- 4
If content is insufficient, call get_document_tree to inspect document structure
- 5
Call get_tree_node on the relevant section node to read full text
Full text of the authentication section with page references and source document metadata
Требования
- Orchata account and API access
Установка
clawhub install ehudsn/orchataВыполните команду CLI или вставьте промпт в Claude Code, Cursor, Codex — любой кодинг-агент, которым пользуетесь.