Run the tool.

Return history buffer by key or all memories.

Adding arguments for running test interactively or setting verbosity

Construct an agent from an LLM and tools.

* Helper function for animation delays, called with `await`. * @param {number} time - Timeout, in ms.

Clear memory contents.

Return history buffer and format it into a conversational string format.

Save context from this conversation to buffer.

Convert tool into function parameter for openai

Create Chain for running tests Args: tools: list of autochain tools memory: memory store for chain llm: model for age

If the message contains a json response, try to parse it into dictionary

Get the key's corresponding value from redis.

Plan the next step. either taking an action with AgentAction or respond to user with AgentFinish Args: history: entire co

Wrapper for _run function by formatting the input and outputs Args: user_query: user query return_only_outputs: boole

* Initialise the widget, get lines, clear container and start animation.

Check if the generation is confident enough to take action

Return history buffer by key or all memories.

Save the context of this model run to memory.

Clear redis memory.

Clear memory contents.

Clear memory contents.

Create LangChainWrapperChain by instantiating LangChain agent Args: tools: list of langchain tool agent_type: LangChain Agent

Use tenacity to retry the completion call.

Get a value from a dictionary or an environment variable.

Create prompt in the style of the zero shot agent. Args: template: message to be injected between prefix and suffix.

Return chat message history.

Return key-value pairs given the text input to the chain.

Return history buffer and format it into a conversational string format.

Return key-value pairs given the text input to the chain.

Save context from this conversation to redis.

Save context from this conversation to buffer.

Save the key value pair to redis.

Save the context of this model run to memory.

Determine if agent should continue to answer user questions based on the latest user query

* Helper function for generating attributes string. * * @param {Object} line - Line data object. * @returns {string} - String of attrib

Cut off the text as soon as any stop words occur.

Convert tool input to pydantic model.

Run inputs including user query and past conversation with agent and get response back calls take_next_step function to determine wha

Add a list of documents to collection

Ask clarifying question if needed. When agent is about to perform an action, we could use this function with different prompt to ask

If the tool failed due to error, what should be the fix for inputs

Get a value from a dictionary or an environment variable.

Estimate the confidence of the generation Args: history: history of the conversation agent_output: the output

* Converts line data objects into line elements. * * @param {Object[]} lineData - Dynamically loaded lines. * @param {Object} line - Li

Parse estimated confidence from the message

Parse the reason and rating from the call to determine if the conversation reaches the expected outcome

Load conversation history from memory and prep inputs.

Save conversation into memory and prep outputs.

* Animate a progress bar. * @param {Node} line - The line element to render.

Run query through GoogleSearch and return metadata. Args: query: The query to search for. num_results: The number of