Function reference • mergen

Setting up an agent

Function for setting up an AI agent

setupAgent(): set up an online LLM API for subsequent tasks

Sending prompts

Functions for sending prompts to your agent

sendPrompt(): Send a prompt to a specified language model agent and return the response.

selfcorrect(): Self correct the code returned by the agent

runCodeInResponse(): Executes the code in the received response from the agent

sampleResponse(): Sample more solutions when non-executable code returned by the agent

Manipulate agent’s answer

Functions for manipulating the agents answer so code can be extracted, executed, filenames can be extracted etc. Some of these functions are internally used by others.

clean_code_blocks(): Clean code blocks returned by the agent

extractCode(): extract the code and text from the text returned by LLM agent

executeCode(): execute code

extractFilenames(): Extract file names from user prompt

fileHeaderPrompt(): Extract file headers from files in prompt

check_install(): Check and Install R Package

extractInstallPkg(): extract package names and install them

promptContext(): Predefined prompt contexts for prompt engineering