Agent as Tool

The simplest multi-agent pattern: wrap a specialised agent as a tool that another agent can call. This requires no new classes — just agent.as_tool().

Basic usage

from elsai import Agent

# A specialised sub-agent
researcher = Agent(
    name="researcher",
    description="Searches the web and returns summarised findings on any topic.",
    system_prompt="You are a research specialist. Find accurate, up-to-date information.",
)

# The orchestrating agent uses the researcher as a tool
orchestrator = Agent(
    name="orchestrator",
    system_prompt="You are a helpful assistant. Use the researcher tool when you need information.",
    tools=[researcher.as_tool()],
)

result = orchestrator("What are the key differences between React and Vue?")
print(result)

as_tool() parameters

researcher.as_tool(
    name="research",             # Tool name (defaults to agent.name)
    description="...",           # Tool description (defaults to agent.description)
    preserve_context=False,      # Whether to keep conversation history between calls
)

Parameter	Default	Description
name	agent.name	Tool name the model sees
description	agent.description	How the model decides when to call this tool
preserve_context	False	True = keep history across calls; False = fresh start each time

Preserve context

Set preserve_context=True when the sub-agent needs to remember earlier interactions:

# Stateful sub-agent — remembers context between calls
memory_agent = Agent(
    name="memory",
    description="Stores and retrieves information from our working memory.",
)

orchestrator = Agent(
    tools=[memory_agent.as_tool(preserve_context=True)],
)

Chains of agents

You can nest agents multiple levels deep:

# Level 3 — data retriever
data_agent = Agent(name="data", description="Retrieves raw data from databases")

# Level 2 — analyst uses data agent
analyst = Agent(
    name="analyst",
    description="Analyses data and produces insights",
    tools=[data_agent.as_tool()],
)

# Level 1 — orchestrator uses analyst
orchestrator = Agent(
    name="orchestrator",
    tools=[analyst.as_tool()],
)

result = orchestrator("Analyse our top customers and explain the key trends")

Passing tools

Sub-agents can have their own tools:

from elsai_tools import calculator, python_repl

coder = Agent(
    name="coder",
    description="Writes and executes Python code to solve problems",
    tools=[calculator, python_repl],
)

manager = Agent(
    name="manager",
    tools=[coder.as_tool()],
)

result = manager("Calculate the compound interest on $10,000 at 5% for 10 years")

Comparison with Graph/Swarm

	as_tool()	Graph	Swarm
Setup	Minimal	Medium	Medium
Orchestration	LLM decides when to call	Developer-defined order	Autonomous
Best for	1–3 specialists	Multi-step pipelines	Open-ended teamwork