Skip to main content
Planning is the agent’s ability to reason about a goal and chart a path of intermediate steps needed to achieve it. Rather than treating every user message or tool call as an isolated request, an agent can generate a structured plan that sequences actions, selects the right tools, and manages dependencies across multiple turns. This makes agents more reliable and efficient for complex tasks, because they can break down a broad objective into smaller, executable parts, adapt their plan when conditions change, and maintain coherence over longer workflows. Planning is a core capability that elevates an agent from a reactive assistant to a proactive, autonomous system that can handle real-world complexity.
images/main/main.py
from autonomy import Agent, Model, Node, ReActPlanner


async def main(node):
    await Agent.start(
        node=node,
        name="henry",
        instructions="You are Henry, an expert legal assistant.",
        model=Model("llama3.2"),
        planner=ReActPlanner(Model("deepseek-r1")),
    )


Node.start(main)
images/main/Dockerfile
FROM ghcr.io/build-trust/autonomy-python
COPY . .
ENTRYPOINT ["python", "main.py"]
autonomy.yaml
name: example007
pods:
  - name: main-pod
    public: true
    containers:
      - name: main
        image: main
⌘I