Split content on structural boundaries. Args: content: Content to split Returns: Li

Split content on structural boundaries. Args: content: Content to split Returns: Li

Transcribe audio file using Whisper ASR via Docling.

Transcribe audio file using Whisper ASR via Docling.

Add the attachments section to the page

Add the code generation tab buttons section to the page

Add the main tab buttons section to the page

Check buttons pushed

Use LLM to generate chunk-specific context

Add documents to a ChromaDB collection in batches. Args: collection: ChromaDB collection ids: List of document IDs do

Add documents to a ChromaDB collection in batches. Args: collection: ChromaDB collection ids: List of document IDs do

Add a new MCP server configuration if the query starts with 'mcpServer'. The query should be in the format: {"server_name": {

Add the models selection to the page

Show the selected conversation's question and answer in the main section

Add the sidebar to the page and return the data management container

Add the suggestions section to the page

Add the title section to the page

Add the user input section to the page and return the question object

Main interaction loop that processes user queries using the LLM and available tools. This function: 1. Sends the user query to the LLM

Perform a Allegro video generation request check

Perform a Allegro video generation request

Analyze and display chunk statistics.

Analyze a matchup and generate prediction.

Analyze NBA matchup with complete data.

Uses LLM to perform comprehensive workflow analysis. Generates three analyses: 1. Overall workflow purpose and functionality 2. Node

Build and return the travel agent graph.

Main chat loop for interacting with Claude. Args: resume_session: If True, attempts to resume the last conversation

Run an interactive chat loop

Check database connection.

Check database connection.

Uses LLM to assess if an n8n workflow is legitimate vs test/spam. Prompts LLM to analyze workflow structure and patterns to determine va

Convert and chunk document using HybridChunker.

Chunk a document into semantically coherent pieces. Args: content: Document content title: Document

Chunk document using simple rules. Args: content: Document content title: Document title

Chunk document using simple rules. Args: content: Document content title: Document title

Split text into chunks, respecting code blocks and paragraphs.

Splits a text into chunks, respecting code blocks, paragraphs, and sentences. This function splits a text into chunks of a specified size, priori

Clean up server resources.

Clean up all resources including the exit stack.

Clean up server resources.

Clean up resources

Clean up all servers properly.

Clear all data from the knowledge graph.

Clear the session for a user (keeps cwd configuration). Args: user_id: Telegram user ID

Close database connections.

Close graph client.

Close database connections.

Close database connections.

Close database connections.

Close database connection pool.

Close database connection pool.

Close database connection pool.

Close database connection pool.

Close database connection pool.

Close database connection pool.

Close database connection pool.

Close database connection pool.

Initialize and connect the Claude SDK client.

Connect to an MCP server using config.json settings

Connect to a server using the provided configuration. Args: server_name (str): The name of the server. serve

Checks if a given string contains a valid YouTube URL. This function uses a regular expression to determine if the input string matches the e

Convert a list of messages to a list of dictionaries.

Convert Claude SDK message stream to OpenAI SSE format. This function takes the async iterator from Claude SDK's receive_messages() and

Convert Claude SDK message stream to OpenAI SSE format. This function takes the async iterator from Claude SDK's receive_messages() and

Crawl content from a list of articles. Args: articles (list): List of articles with URLs session_id (str): Session ID fo

Batch crawl using logic from 3-crawl_sitemap_in_parallel.py.

Crawl a .txt or markdown file using logic from 4-crawl_and_chunk_markdown.py.

Crawl multiple URLs in parallel with a concurrency limit.

Recursive crawl using logic from 5-crawl_recursive_internal_links.py. Returns list of dicts with url and markdown.

Fetch and parse content from a URL. Args: url (str): URL to crawl session_id (str): Session ID for logging Retu

Factory method to create specialized agents for each step

Create appropriate chunker based on configuration. Args: config: Chunking configuration Returns: Chunker instan

Create appropriate chunker based on configuration. Args: config: Chunking configuration Returns: Chunker instan

Create a dynamic tool for a given server and tool. Args: tool: The tool object. server_name (str): The name

Creates a knowledge base from a user-uploaded document. This function handles the creation of a knowledge base from a document, including ret

Convert MCP tools to pydantic_ai Tools.

Convert MCP tools to pydantic_ai Tools.

Convert MCP tools to pydantic_ai Tools.

Create FunctionModel that simulates search behavior. Args: search_results: Expected search results to return Returns:

Create a self-signed certificate for HTTPS

Initialize a Pydantic AI Tool from an MCP Tool.

Initialize a Pydantic AI Tool from an MCP Tool.

Initialize a Pydantic AI Tool from an MCP Tool.

Create and persist Chroma vector store.

Show data management components in the side bar

Deactivates all streams and marks all buzz as inactive for a given session. This function deactivates streams associated with the session, and up