Process a non-streaming completion response from OpenRouter.

Process a streaming response from OpenRouter.

Async call to Google AI embedding API. Note: Google AI Python client doesn't have async support yet, so this falls back to sy

kwargs is the combined input and model_kwargs

Make an asynchronous call to the AWS Bedrock API.

Make an asynchronous call to the OpenRouter API.

Safely append a dialog turn to the conversation

kwargs is the combined input and model_kwargs. Support streaming call.

Return the conversation history as a dictionary.

Process a query using RAG. Args: query: The user's query Returns: Tuple of (RAGAnswer, retrieved_do

kwargs is the combined input and model_kwargs. Support streaming call.

Stream a chat completion response directly using Google Generative AI

r""" Specify the API input type and output api_kwargs that will be used in _call and _acall methods. Convert the Component's standard

Convert inputs to API kwargs for AWS Bedrock.

Convert AdalFlow inputs to OpenRouter API format.

Deletes a specific wiki cache from the file system.

Estimate the token count of a given text. Args: text (str): The text to estimate token count for. Returns: int: Estimat

Export wiki content as Markdown or JSON. Args: request: The export request containing wiki pages and format Returns: A

Create an instance from a dictionary.

Create an instance from a dictionary.

r"""When the n > 1, get all the messages content.

r"""When the n > 1, get all the messages content.

Check if authentication is required for the wiki.

Retrieves cached wiki data (structure and generated pages) for a repository.

When we only need the content of the first message.

r"""When we only need the content of the first message. It is the default parser for chat completion.

r"""When we only need the content of the first message. It is the default parser for chat completion.

Return the file tree and README content for a local repository.

Get available model providers and their models. This endpoint returns the configuration of available model providers and their respectiv

r"""Get the probabilities of each token in the completion.

r"""Get the probabilities of each token in the completion.

Lists all processed projects found in the wiki cache directory. Projects are identified by files named like: deepwiki_cache_{repo_type}_{owne

r"""Handle the streaming response.

r"""Handle the streaming response.

Handle WebSocket connection for chat completions. This replaces the HTTP streaming endpoint with a WebSocket connection.

Health check endpoint for Docker and monitoring

Initialize asynchronous LiteLLM OpenAI-compatible client.

Initialize the asynchronous AWS Bedrock client. Note: boto3 doesn't have native async support, so we'll use the sync client i

Initialize synchronous LiteLLM OpenAI-compatible client.

Parse the completion, and put it into the raw_response.

Parse the completion, and put it into the raw_response.

r"""Parse the embedding response to a structure AdalFlow components can understand. Should be called in ``Embedder``.

Parse Bedrock embedding response to EmbedderOutput format.

r"""Parse the embedding response to a structure Adalflow components can understand. Should be called in ``Embedder``.

Parse the image generation response into a GeneratorOutput.

Prepare the retriever for a repository. This is a compatibility method for the isolated API. Args: repo_type(str

Root endpoint to check if the API is running and list available endpoints dynamically.

Root endpoint to check if the API is running

Run a test function with multiple parameter sets.

Set up test fixtures before each test method.

Stores generated wiki data (structure and pages) to the server-side cache.

Test binary assumptions in data pipeline functions.

Test that RAG doesn't make binary assumptions about embedders.

Test that configurations load properly.

Test that all embedder configurations load properly.

Test embedder selection mechanism.