Module: Phronomy::Agent::Concerns::Guardrailable

Included in:
Base
Defined in:
lib/phronomy/agent/concerns/guardrailable.rb

Overview

Adds input and output guardrail support to an agent.

Included in Base. Guardrails are run on the raw input string before the LLM is called, and on the raw output string before the result is returned to the caller.

Instance Method Summary collapse

Instance Method Details

#add_input_guardrail(guardrail) ⇒ self

Attach a guardrail that validates input before every #invoke call.

Parameters:

Returns:

  • (self) 


15
16
17
18
19
 
# File 'lib/phronomy/agent/concerns/guardrailable.rb', line 15

def add_input_guardrail(guardrail)
  @input_guardrails ||= []
  @input_guardrails << guardrail
  self
end

#add_output_guardrail(guardrail) ⇒ self

Attach a guardrail that validates output before it is returned.

Parameters:

Returns:

  • (self) 


24
25
26
27
28
 
# File 'lib/phronomy/agent/concerns/guardrailable.rb', line 24

def add_output_guardrail(guardrail)
  @output_guardrails ||= []
  @output_guardrails << guardrail
  self
end