Package index • tidyprompt

Prompt wrapping and evaluation

Tidyprompt

Functions to create and handle tidyprompt objects.

tidyprompt(): Create a tidyprompt object

tidyprompt-class: Tidyprompt R6 Class

is_tidyprompt(): Check if object is a tidyprompt object

construct_prompt_text(): Construct prompt text from a tidyprompt object

get_prompt_wraps(): Get prompt wraps from a tidyprompt object

get_chat_history(): Get the chat history of a tidyprompt object

set_chat_history(): Set the chat history of a tidyprompt object

set_system_prompt(): Set system prompt of a tidyprompt object

Prompt wrap

Functions for creating prompt wraps.

prompt_wrap(): Wrap a prompt with functions for modification and handling the LLM response

llm_feedback(): Create an llm_feedback object

llm_break(): Create an llm_break object

Send prompt

Function to evaluate prompts with an LLM provider.

send_prompt(): Send a prompt to a LLM provider

Prompt wrap library

A library of pre-built prompt wraps for common use cases.

Answer as

Functions to have LLMs answer in a specific format (structured output).

answer_as_boolean(): Make LLM answer as a boolean (TRUE or FALSE)

answer_as_integer(): Make LLM answer as an integer (between min and max)

answer_as_json(): Make LLM answer as JSON (with optional schema)

answer_as_key_value(): Make LLM answer as a list of key-value pairs

answer_as_list(): Make LLM answer as a list of items

answer_as_named_list(): Make LLM answer as a named list

answer_as_regex_match(): Make LLM answer match a specific regex

answer_as_text(): Make LLM answer as a constrained text response

Answer by

Functions to have LLMs answer in a specific way.

answer_by_chain_of_thought(): Set chain of thought mode for a prompt

answer_by_react(): Set ReAct mode for a prompt

Answer using

Functions that give LLMs access to tools (function-calling) and code

answer_using_r(): Enable LLM to draft and execute R code

answer_using_sql(): Enable LLM to draft and execute SQL queries on a database

answer_using_tools(): Enable LLM to call R functions

tools_add_docs(): Add tidyprompt function documentation to a function

tools_get_docs(): Extract documentation from a function

Miscellaneous

add_text(): Add text to a tidyprompt

quit_if(): Make evaluation of a prompt stop if LLM gives a specific response

user_verify(): Have user check the result of a prompt (human-in-the-loop)

llm_verify(): Have LLM check the result of a prompt (LLM-in-the-loop)

LLM providers & chat history

Functions to interact with LLM providers and manage chat history.

LLM provider class

llm_provider-class: LlmProvider R6 Class

Pre-built LLM providers

llm_provider_google_gemini(): Create a new Google Gemini LLM provider

llm_provider_groq(): Create a new Groq LLM provider

llm_provider_mistral(): Create a new Mistral LLM provider

llm_provider_ollama(): Create a new Ollama LLM provider

llm_provider_openai(): Create a new OpenAI LLM provider

llm_provider_openrouter(): Create a new OpenRouter LLM provider

llm_provider_xai(): Create a new XAI (Grok) LLM provider

Chat history

chat_history(): Create or validate chat_history object

add_msg_to_chat_history(): Add a message to a chat history

Persistent chat

A helper class to have a persistent, manual conversation with a LLM provider.

persistent_chat-class: PersistentChat R6 class

Helper functions

Text helpers

Functions to create text to be used in prompts.

df_to_string(): Convert a dataframe to a string representation

vector_list_to_string(): Convert a named or unnamed list/vector to a string representation

skim_with_labels_and_levels(): Skim a dataframe and include labels and levels

Other helper functions

extract_from_return_list(): Function to extract a specific element from a list

r_json_schema_to_example(): Generate an example object from a JSON schema