Main function that sets up the MCP server, initializes tools, and runs the interactive loop.

Main function that sets up the MCP server, initializes tools, and runs the interactive loop. The server is run in a Docker container to ensur

Parse date from query with timezone handling.

Parse date from query with timezone handling.

Creates the output directory if it doesn't exist

Phase 1: Add episodes about current top LLMs.

Phase 2: Claude 4 emerges as the new best LLM.

Phase 3: MLMs make LLMs irrelevant.

Populate a StreamMetadata object with video details from a snippet. This function extracts relevant information such as the title, channel title,

Post a tweet using the Twitter API with OAuth 1.0a. Args: tweet_text (str): The text to tweet session_id (str): Session

Create domain-specific query-document pairs

Print welcome banner.

Print welcome banner.

Print welcome banner.

Print help information.

Print help information.

Print help information.

Print conversation statistics.

Print conversation statistics.

Processes all active streams. This function iterates through a list of active streams, fetches the latest chat messages for each stream

Process a document and store its chunks in parallel.

Processes a document by splitting it into chunks, extracting metadata, and storing it. This function orchestrates the processing of a document by

Processes buzzes that are in the 'FOUND' state. This function retrieves buzzes in the 'FOUND' state from the database, updates their sta

Processes a list of chat messages for a given session. This function iterates through a list of chat messages, classifies the intent of

Process a single chunk of text.

Processes a single chunk of text to extract metadata and generate embeddings. This function orchestrates the extraction of title and summary usin

Process a single document and return metadata.

Process a document file and store it in the knowledge base. Args: file_path: Path to the document file Returns:

Process a single file with caching.

Process the ideation form

Process the ideation form

Generates the JSON file and GS python code for Tools

Generates the JSON file and GS python code for Tools

Process No-Form buttons, like the ones for the the ideation-from-prompt feature

Main function to process the latest video from a playlist.

Process the Pydantic AI documentation. Args: chunk_size: The size of each chunk overlap: The overlap between chunks

Process a query using Claude and available tools

Orchestrate the entire workflow to get the final summary

Process the use_response_as_prompt button pushed

Process a single video by ID.

Converts n8n workflow data into an HTML component string. Takes raw workflow data, escapes special characters, and wraps in n8n-demo

Store embedding in cache.

Store embedding in cache.

Store embedding in cache.

Store embedding in cache.

Query a ChromaDB collection for similar documents. Args: collection: ChromaDB collection query_text: Text to search for

Query a ChromaDB collection for similar documents. Args: collection: ChromaDB collection query_text: Text to search for

Decide where to read arguments from

Read arguments from the command line

Reads a JSON file and returns its content as a dictionary

Reads and processes live chat messages from active YouTube streams. This function retrieves all active stream sessions, fetches live chat

Run the agent with streaming text for the user_input prompt, while maintaining the entire conversation in `st.session_state.messages`.

Run the RAG agent with streaming response. Args: user_input: User query Yields: Streamed response chunks

Run all integration tests.

Run the agent in an interactive CLI with streaming.

Run the agent in an interactive CLI with streaming.

Run the agent in an interactive CLI with streaming.

Run the RAG agent to answer a question about Pydantic AI. Args: question: The question to answer. Returns: The a

Run the RAG agent to answer a question about Pydantic AI. Args: question: The question to answer. collection_name: Name of th

Run the RAG agent to answer a question about Pydantic AI. Args: question: The question to answer. collection_name: Name of th

Run the travel agent with the given user input.

Save chunks to file with separators, preserving context and headings.

Saves the file to the output directory

Save the item in the database

Save the session ID to a file for resuming later.

Save session data for a Telegram user. Args: user_id: Telegram user ID session_id: Claude SDK session ID cwd: Workin

Scrape a Markdown page and split into chunks by # and ## headers.

Search for document chunks by vector similarity. Args: query_embedding: Vector embedding of the query

Search the knowledge graph. Args: query: Search query Returns: Search results

Chunk based on semantic similarity, not fixed size

Send a long message, splitting it if necessary due to Telegram's 4096 character limit. Args: chat_id: Telegram chat ID text: