Chaining

Run agents in sequence. Output of each agent feeds into the next.

Basic usage

from pure_agents import Agent, chain

researcher = Agent(template="researcher")
summariser = Agent(system_prompt="Summarise the following text concisely.")
formatter = Agent(system_prompt="Format as bullet points.")

result = await chain(
    [researcher, summariser, formatter],
    "The history of machine learning"
)

How it works

prompt → Agent 1 → output → Agent 2 → output → Agent 3 → final result

Each agent receives the previous agent's output as its prompt.

Use cases

Research pipeline:

gatherer = Agent(system_prompt="Find information about the topic.")
analyser = Agent(system_prompt="Analyse the information critically.")
writer = Agent(system_prompt="Write a report based on the analysis.")

report = await chain([gatherer, analyser, writer], "Climate change in 2024")

Content pipeline:

drafter = Agent(system_prompt="Write a first draft.")
editor = Agent(system_prompt="Improve clarity and fix errors.")
formatter = Agent(system_prompt="Format for publication.")

article = await chain([drafter, editor, formatter], "AI in healthcare")

Code pipeline:

planner = Agent(system_prompt="Plan the implementation approach.")
coder = Agent(template="coder")
reviewer = Agent(system_prompt="Review the code for bugs and improvements.")

code = await chain([planner, coder, reviewer], "REST API for user management")

Sync version

from pure_agents import chain_sync

result = chain_sync([agent1, agent2], "Your prompt")

Tips

• Keep each agent focused on one task
• Order matters: think about what each agent needs as input
• Use debug=True on agents to see intermediate outputs