Skip to content


6. Detailed Module API

agenticaiframework.agents

  • Agent – Base class for all agents.
  • AgentManager – Manages multiple agents.

agenticaiframework.communication

  • send_message() – Send a message to another agent or system.
  • receive_message() – Receive a message.

agenticaiframework.configurations

  • set_config(key, value) – Set a configuration value.
  • get_config(key) – Retrieve a configuration value.

agenticaiframework.evaluation

  • evaluate_response(response) – Evaluate the quality of a response.

agenticaiframework.guardrails

  • add_guardrail(func) – Add a guardrail function.
  • remove_guardrail(func) – Remove a guardrail.

agenticaiframework.hub

  • register_agent(name, cls) – Register an agent.
  • get_agent(name) – Retrieve an agent.
  • register_tool(name, func) – Register a tool.
  • get_tool(name) – Retrieve a tool.

agenticaiframework.knowledge

  • KnowledgeBase – Manage documents and search.

agenticaiframework.llms

  • OpenAIModel – Example LLM integration.

agenticaiframework.mcp_tools

  • load_tool(name) – Load an MCP tool.

agenticaiframework.memory

  • Memory – Store and retrieve key-value pairs.

agenticaiframework.monitoring

  • log_event(message, level) – Log an event.

agenticaiframework.processes

  • run_process(name, params) – Run a process.
  • run_process_async(func) – Run a process asynchronously.

agenticaiframework.prompts

  • PromptTemplate – Manage prompt templates.

agenticaiframework.tasks

  • Task – Define and manage tasks.

7. Advanced Usage Examples

Using Multiple Modules Together

from agenticaiframework import Agent
from agenticaiframework.hub import register_agent, get_agent, register_tool
from agenticaiframework.memory import Memory

class EchoAgent(Agent):
    def act(self, input_data):
        return f"Echo: {input_data}"

register_agent("echo", EchoAgent)

memory = Memory()
memory.store("greeting", "Hello")

agent = get_agent("echo")
print(agent.act(memory.retrieve("greeting")))

Custom Process with Guardrails

from agenticaiframework.guardrails import add_guardrail
from agenticaiframework.processes import run_process

def no_numbers(input_data):
    if any(char.isdigit() for char in input_data):
        raise ValueError("Numbers are not allowed!")
    return input_data

add_guardrail(no_numbers)

def greet():
    return "Hello, World!"

print(run_process(greet))

8. Notes on API Stability

  • Public APIs follow semantic versioning.
  • Experimental APIs are marked in the documentation and may change.

AgenticAI API Reference

This document provides a reference for the main classes, functions, and modules in AgenticAI.

1. Modules Overview

  • agenticaiframework.agents – Agent base classes and implementations.
  • agenticaiframework.communication – Communication utilities.
  • agenticaiframework.configurations – Configuration management.
  • agenticaiframework.evaluation – Evaluation and scoring.
  • agenticaiframework.guardrails – Safety and compliance checks.
  • agenticaiframework.hub – Registry for agents, tools, and processes.
  • agenticaiframework.knowledge – Knowledge base management.
  • agenticaiframework.llms – LLM integrations.
  • agenticaiframework.mcp_tools – MCP tool integrations.
  • agenticaiframework.memory – Memory management.
  • agenticaiframework.monitoring – Monitoring and logging.
  • agenticaiframework.processes – Workflow orchestration.
  • agenticaiframework.prompts – Prompt templates.
  • agenticaiframework.tasks – Task management.

2. Core Classes

Agent

Location: agenticaiframework.agents

Methods: - act(input_data) – Perform an action based on input. - observe(data) – Observe environment or input. - plan() – Plan next steps.

Memory

Location: agenticaiframework.memory

Methods: - store(key, value) – Store a value. - retrieve(key) – Retrieve a value. - clear() – Clear memory.

Hub

Location: agenticaiframework.hub

Functions: - register_agent(name, cls) – Register an agent. - get_agent(name) – Retrieve an agent instance. - register_tool(name, func) – Register a tool. - get_tool(name) – Retrieve a tool.

3. Utility Functions

set_config(key, value)

Location: agenticaiframework.configurations

Set a configuration value.

run_process(name, params)

Location: agenticaiframework.processes

Run a registered process.

4. Example Usage

from agenticaiframework.hub import get_agent

agent = get_agent("default_agent")
print(agent.act("Hello"))

5. Notes

  • All public APIs are subject to semantic versioning.
  • Internal APIs may change without notice.