Vedarama

Search Vedic Scriptures

Search across Vedic scriptures library using semantic vector search. Best for finding passages by topic or meaning. Returns relevant text chunks with book, section, language, and relevance score. === WHEN TO USE === Use this tool when user asks a QUESTION about a topic (not a specific verse reference). Examples: - "что говорится о карме?" → search_vedic_scriptures(query="карма") - "what is the nature of the soul?" → search_vedic_scriptures(query="nature of the soul") - "Что Баладева Видьябхушан

Get Book Chunks

Search within a specific Vedic scripture by its ID, with optional chapter/verse filtering. === WHEN TO USE === Use this tool when user explicitly asks to search within a specific book. Algorithm: 1) find_book_semantic to get book_id (find_book_exact only as fallback), 2) get_book_chunks(book_id, query, ...) === FILTERING === - chapter_filter: filter by chapter/section name (e.g. "глава 4", "canto 3", "Madhya") - verse_range: filter by verse numbers {start: 30, end: 40}

Search Exact

SUPPLEMENTARY tool: Search for exact text matches in the raw book library. Use ONLY when user explicitly asks to find an exact quote, specific Sanskrit term, or precise phrase. Prefer search_vedic_scriptures (semantic search) as the PRIMARY content search tool. Use search_exact as a fallback when semantic search misses a specific string, or when user says "find exact...", "find the phrase...". IMPORTANT: For performance, always provide book_ids to limit the search scope.

Find Book Semantic

PRIMARY tool for finding books by name, author, or description. Handles transliteration variants (Goswami/Gosvami), fuzzy queries, and synonyms. ALWAYS use this tool FIRST for any book lookup. Include author name in query string instead of author_name parameter. Example: find_book_semantic(query="Bhakti-sandarbha Jiva Gosvami") instead of separate author_name. Response includes: book ID, slug, authors, commentators, languages, score.

Find Book Exact

Exact substring search for books by name, slug, or author. SUPPLEMENTARY tool — use only as FALLBACK when find_book_semantic returned no results, or when user explicitly needs exact string matching. Prefer find_book_semantic for all book lookups — it handles transliteration variants and fuzzy matches. Response includes: book ID, slug, authors, commentators, languages, score.

Find Topic Semantic

PRIMARY tool for finding books by topic or philosophical concept. Searches by meaning in book descriptions. Use for descriptive queries: "books about karma", "philosophy of devotion". ALWAYS use this tool FIRST for topic-based book lookups. Use find_topic_exact only as fallback. Response includes: book ID, authors, commentators, description, score.

Find Topic Exact

SUPPLEMENTARY tool: Exact substring search for books by topic — searches in categories and names. Use ONLY as fallback when find_topic_semantic returned no results, or when user needs exact term matching. Prefer find_topic_semantic as the PRIMARY topic search tool. Response includes: book ID, authors, commentators, description, score.

Get Embedding Status

Get detailed embedding status for books in the library. Shows which books have been embedded into the vector database and which have not. For books without embeddings, shows the reason: no chunk files, embedding failed, upload failed, etc. Use this to diagnose why certain books are not searchable via semantic search.

Get Description

Get book annotation and/or chapter descriptions (title, summary) for a specific book. Use this tool to: - Get a book's annotation/description by book_id or book_slug - Get the list of all chapters (table of contents) of a book - Get the detailed summary of a specific chapter For main scriptures (SB, BG, CC, CB), you can also pass a short reference like "SB 8" or "BG" to get the chapter list, or "SB 8.2" to get chapter 2 of canto 8 summary. Algorithm: 1) To get a book description: pass

Get Verse

IMPORTANT: You MUST use this tool whenever the user mentions specific verse numbers or references. Always call this tool FIRST to retrieve the actual verse text before answering. Look up a specific verse with all available commentaries from the library on disk. Searches across all books that cover the same text (e.g. all Bhagavad-gita commentaries). Returns: sanskrit, transliteration, word-by-word, translation, and commentary for each book. === ALGORITHM for finding verse text === 1. For main

Vedarama Description

Returns a detailed description of the Vedarama.ru platform. Use when user asks about Vedarama.ru, the library, or where the texts come from.

Vedarama Chat Description

Returns a detailed description of the Vedarama AI application. Use when user asks about the chat features, capabilities, or how to get access.