Agent Spec Adapters - WayFlow

↑ With the Agent Spec adapter for WayFlow, you can easily import agents from external frameworks using Agent Spec and run them with WayFlow.

WayFlow is the reference framework for Agent Spec, provides modular components for developing AI-powered assistants, supporting both workflow-based and agent-style applications.

Get started

To get started, set up your Python environment (Python 3.10 or newer required), and then install the PyAgentSpec package as well as WayFlowCore.

python -m venv .venv
source .venv/bin/activate  # On Windows: .venv\Scripts\activate
pip install pyagentspec wayflowcore

You are now ready to use the adapter:

• Run Agent Spec configurations with WayFlow (see more details below)

• Convert WayFlow agents to Agent Spec (see more details below)

Run Agent Spec configurations with WayFlow

# Create a Agent Spec agent
from pyagentspec.agent import Agent
from pyagentspec.llms.openaicompatibleconfig import OpenAiCompatibleConfig
from pyagentspec.property import FloatProperty
from pyagentspec.tools import ServerTool

subtraction_tool = ServerTool(
    name="subtraction-tool",
    description="subtract two numbers together",
    inputs=[FloatProperty(title="a"), FloatProperty(title="b")],
    outputs=[FloatProperty(title="difference")],
)

agentspec_llm_config = OpenAiCompatibleConfig(
    name="llama-3.3-70b-instruct",
    model_id="/storage/models/Llama-3.3-70B-Instruct",
    url="url.to.my.llm",
)

agentspec_agent = Agent(
    name="agentspec_tools_test",
    description="agentspec_tools_test",
    llm_config=agentspec_llm_config,
    system_prompt="Perform subtraction with the given tool.",
    tools=[subtraction_tool],
)

# Export the Agent Spec configuration
from pyagentspec.serialization import AgentSpecSerializer

agentspec_config = AgentSpecSerializer().to_json(agentspec_agent)

# Load and run the Agent Spec configuration with WayFlow
from wayflowcore.agentspec import AgentSpecLoader

def subtract(a: float, b: float) -> float:
    return a - b

async def main():
    converter = AgentSpecLoader(tool_registry={"subtraction-tool": subtract})
    assistant = converter.load_json(agentspec_config)
    conversation = assistant.start_conversation()

    while True:
        user_input = input("USER >> ")
        if user_input == "exit":
            break
        conversation.append_user_message(user_input)
        await conversation.execute_async()
        last = conversation.get_last_message()
        print(f"AGENT >> {last.content}")

# anyio.run(main)
# USER >> Compute 987654321-123456789
# AGENT >> The result of the subtraction is 864197532.

Convert WayFlow agents to Agent Spec

# Create a WayFlow Agent
from wayflowcore.agent import Agent
from wayflowcore.models import OpenAICompatibleModel
from wayflowcore.tools import tool

@tool("subtraction-tool", description_mode="only_docstring")
def subtraction_tool(a: float, b: float) -> float:
    """subtract two numbers together"""
    return a - b

llm = OpenAICompatibleModel(
    name="llama-3.3-70b-instruct",
    model_id="/storage/models/Llama-3.3-70B-Instruct",
    base_url="url.to.my.llm",
)

wayflow_agent = Agent(
    name="wayflow_agent",
    description="Simple agent with a tool.",
    llm=llm,
    custom_instruction="Perform subtraction with the given tool.",
    tools=[subtraction_tool],
)

# Convert to Agent Spec
from wayflowcore.agentspec import AgentSpecExporter

agentspec_config = AgentSpecExporter().to_json(wayflow_agent)