Skip to main content

NPCs (Non-Player Characters)

NPCs in PatientHub are supporting agents that can participate in therapy simulations beyond the primary client-therapist dyad.

Available NPCs​

NPCKeyDescription
InterviewerinterviewerConducts structured scripted interviews

Usage​

In Code​

from omegaconf import OmegaConf
from patienthub.npcs.interviewer import InterviewerNPC

config = OmegaConf.create({"data": "data/evaluations/surveys/default_survey.json"})
interviewer = InterviewerNPC(configs=config)

question = interviewer.generate_response("start")
print(question)

Use Cases​

Intake Interview (Before Therapy)​

from omegaconf import OmegaConf
from patienthub.clients import get_client
from patienthub.therapists import get_therapist
from patienthub.npcs.interviewer import InterviewerNPC

# Create interviewer NPC (question script)
interviewer = InterviewerNPC(
    configs=OmegaConf.create({"data": "data/evaluations/surveys/default_survey.json"})
)

# Create client
client = get_client(agent_name="patientPsi", lang="en")

# Create therapist
therapist = get_therapist(agent_name="basic", lang="en")

# Run a quick intake, then start therapy
question = interviewer.generate_response("start")
print("Interviewer:", question)
answer = client.generate_response(question)
print("Client:", answer.content if hasattr(answer, "content") else answer)

Creating Custom NPCs​

To add a new NPC, create a new class under patienthub/npcs/ and instantiate it directly. NPCs are lightweight — simply implement generate_response() and reset().

from omegaconf import DictConfig

class FamilyMemberNPC:
    def __init__(self, configs: DictConfig):
        self.configs = configs
        self.name = "Family Member"
        self.reset()

    def generate_response(self, msg: str):
        return "I'm here to support them."

    def reset(self):
        self.messages = []

See Also​