Python SDK Overview

Full API parity with the TypeScript core — snake_case, Pydantic validation, async/await.

Installation

pip install agentforge

API Parity

The Python SDK mirrors the TypeScript API with Python conventions:

TypeScript	Python
`camelCase`	`snake_case`
`Zod` schemas	`Pydantic` models
`Promise<T>`	`async/await`
`AsyncGenerator`	`async for`

Quick Example

import { defineAgent, defineTool, createLLM } from '@ahzan-agentforge/core';
import { z } from 'zod';

const tool = defineTool({
  name: 'greet',
  description: 'Greet someone',
  input: z.object({ name: z.string() }),
  output: z.object({ message: z.string() }),
  execute: async ({ name }) => ({ message: `Hello, ${name}!` }),
});

const agent = defineAgent({
  name: 'greeter',
  description: 'Greets people',
  tools: [tool],
  llm: createLLM({ provider: 'anthropic', model: 'claude-sonnet-4-20250514' }),
  systemPrompt: 'You greet people warmly.',
});

from agentforge import define_agent, define_tool, create_llm
from pydantic import BaseModel

class GreetInput(BaseModel):
    name: str

class GreetOutput(BaseModel):
    message: str

tool = define_tool(
    name="greet",
    description="Greet someone",
    input=GreetInput,
    output=GreetOutput,
    execute=lambda inp: GreetOutput(message=f"Hello, {inp.name}!"),
)

agent = define_agent(
    name="greeter",
    description="Greets people",
    tools=[tool],
    llm=create_llm(provider="anthropic", model="claude-sonnet-4-20250514"),
    system_prompt="You greet people warmly.",
)

What's Included

The Python SDK includes everything from the TypeScript core:

Agent, define_agent, define_tool, create_llm
All state stores, memory stores, and testing utilities
Budget governor, autonomy policy, rollback
Multi-agent coordination (Coordinator, pipeline, parallel, supervisor, debate)
Error types, tracing, and observability

Next Steps

Quickstart — Python-specific setup
API Parity — complete mapping table
Differences — Python-specific differences