Agent LLM Node R6 Class

A specialized AgentNode that leverages a Large Language Model (LLM) to generate outputs from prompts. It manages prompt construction by combining a persistent role (system prompt) with dynamic context from the AgentState. It also handles tool injection and automatic context discovery from local files (e.g., agents.md, skills.md).

Value

An AgentLLMNode object.

Super class

HydraR::AgentNode -> AgentLLMNode

Public fields

role

String. System prompt/role for the agent.

model

String. Default model.

driver

AgentDriver object.

output_format

String. Output expectation.

cli_opts

List. Default CLI options for the driver.

prompt_builder

Function(state) -> String.

tools

List of AgentTool objects.

agents_files

Character vector. Paths to agents context files.

skills_files

Character vector. Paths to skills context files.

Methods

Public methods

• AgentLLMNode$new()

• AgentLLMNode$run()

• AgentLLMNode$swap_driver()

• AgentLLMNode$clone()

---

Method new()

Initialize AgentLLMNode

Usage

AgentLLMNode$new(
  id,
  role,
  driver,
  model = NULL,
  cli_opts = list(),
  prompt_builder = NULL,
  tools = list(),
  label = NULL,
  params = list(),
  agents_files = NULL,
  skills_files = NULL
)

Arguments

id

String. Unique identifier for the node.

role

String. The primary system prompt or persona the LLM should assume.

driver

AgentDriver. An instance of an AgentDriver subclass (CLI or API based).

model

String. Optional. The specific model to use (overrides driver default).

cli_opts

List. Optional. Named list of parameters for the LLM call (e.g., temperature).

prompt_builder

Function. Optional. A function that takes an AgentState and returns a string prompt. If omitted, the node serializes the entire state as JSON.

tools

List. A list of AgentTool objects available for the agent to use.

label

String. Human-readable name for visualization.

params

List. Additional configuration (e.g., output_format="r").

agents_files

Character vector. Optional paths to markdown files containing agent interaction guidelines.

skills_files

Character vector. Optional paths to markdown files containing specialized tool instructions. Run the LLM Node

---

Method run()

Executes the LLM call. This method handles prompt construction, tool injection, context file discovery, and driver invocation.

Usage

AgentLLMNode$run(state, ...)

Arguments

state

AgentState. The centralized state object for the workflow.

...

Additional arguments. Passed through to the driver's call() method.

Returns

A list containing status, output (the LLM response), raw (the full driver response), and meta-information. Swap Driver at Runtime

---

Method swap_driver()

Usage

AgentLLMNode$swap_driver(driver)

Arguments

driver

AgentDriver object or String ID.

---

Method clone()

The objects of this class are cloneable with this method.

Usage

AgentLLMNode$clone(deep = FALSE)

Arguments

deep

Whether to make a deep clone.

Examples

if (FALSE) { # \dontrun{
# NOTE: Set ANTHROPIC_API_KEY in your .Renviron file
driver <- AnthropicAPIDriver$new()

# Create a node with a prompt builder that pulls from state
node <- AgentLLMNode$new(
  id = "summarizer",
  role = "You are a concise summarizer.",
  driver = driver,
  prompt_builder = function(state) {
    sprintf("Summarise this text: %s", state$get("input_text"))
  }
)
} # }