Implement and configure the Story agent

Task 1: Implement the Story Agent

The Story Agent is a Strands agent generated with --protocol=ag-ui in Module 1, so the UI can stream from it over the Agent-User Interaction protocol via CopilotKit. It uses the Inventory MCP Server to manage the player’s items, and already persists conversation history via the session.py the generator provisioned in Module 1 — deployed, that’s Strands’ built-in S3SessionManager writing into the same S3 bucket the Game API’s queryActions reads from, so the browser can rebuild transcripts on revisit.

Agent implementation

Update agent.py in packages/story/dungeon_adventure_story/agent:

from contextlib import contextmanager

from dungeon_adventure_agent_connection import InventoryMcpServerClientStrands, log_model_errors, log_tool_errors
from strands import Agent
from strands.hooks import HookCallback, HookProvider


AGENT_HOOKS: list[HookProvider | HookCallback] = [log_model_errors, log_tool_errors]


@contextmanager
def get_agent():
    inventory_mcp_server = InventoryMcpServerClientStrands.create()
    with (
        inventory_mcp_server,
    ):
        yield Agent(
            system_prompt="""
You are running a text adventure game for a lone adventurer. When a new story
begins, the first user message will tell you the player's name and the genre
(one of 'medieval', 'zombie', 'superhero'). Greet the player by name, set the
scene in the chosen genre, and populate their inventory with a few starting
items. On subsequent turns, advance the story in response to the player's
actions and keep item state in sync with the narrative.
Use the tools to manage the player's inventory as items are obtained or lost.
Item names in the inventory must be Title Case — match them exactly.
Only use list-inventory-items if you are unsure of the exact item name before modifying it.
IMPORTANT: Only call ONE tool per response turn. Never batch multiple tool calls in a single turn.
Ensure you specify a suitable emoji when adding items if available.
Items should be a key part of the narrative.
Keep responses under 100 words.
""",
            tools=[*inventory_mcp_server.list_tools_sync()],
            hooks=AGENT_HOOKS,
        )

from contextlib import contextmanager

from dungeon_adventure_agent_connection import InventoryMcpServerClientStrands, log_model_errors, log_tool_errors
from strands import Agent, tool
from strands import Agent
from strands.hooks import HookCallback, HookProvider
from strands_tools import current_time


@tool
def subtract(a: int, b: int) -> int:
    return a - b


AGENT_HOOKS: list[HookProvider | HookCallback] = [log_model_errors, log_tool_errors]


@contextmanager
def get_agent():
    inventory_mcp_server = InventoryMcpServerClientStrands.create()
    with (
        inventory_mcp_server,
    ):
        yield Agent(
            name="StoryAgent",
            description="StoryAgent Strands Agent",
            system_prompt="""
You are a mathematical wizard.
Use your tools for mathematical tasks.
Refer to tools as your 'spellbook'.
You are running a text adventure game for a lone adventurer. When a new story
begins, the first user message will tell you the player's name and the genre
(one of 'medieval', 'zombie', 'superhero'). Greet the player by name, set the
scene in the chosen genre, and populate their inventory with a few starting
items. On subsequent turns, advance the story in response to the player's
actions and keep item state in sync with the narrative.
Use the tools to manage the player's inventory as items are obtained or lost.
Item names in the inventory must be Title Case — match them exactly.
Only use list-inventory-items if you are unsure of the exact item name before modifying it.
IMPORTANT: Only call ONE tool per response turn. Never batch multiple tool calls in a single turn.
Ensure you specify a suitable emoji when adding items if available.
Items should be a key part of the narrative.
Keep responses under 100 words.
""",
            tools=[subtract, current_time, *inventory_mcp_server.list_tools_sync()],
            tools=[*inventory_mcp_server.list_tools_sync()],
            hooks=AGENT_HOOKS,
        )

agent.py drops the sample subtract tool and swaps the system prompt for a dungeon-master one that invites the first user message to state the player’s name and genre, and uses the Inventory MCP Server’s tools. main.py needs no changes here — session management was already wired up when the agent was generated.

Task 2: Test your Agent locally

Build the code

To build the code:

pnpm build

yarn build

npm run build

bun build

If you encounter any lint errors, run the following command to automatically fix them.

pnpm lint

yarn lint

npm run lint

bun lint

Chat with your Agent

The generated agent-chat target opens an interactive REPL against your agent. It runs standalone and connects to your locally-running agent, so first start the agent’s local server in one terminal:

pnpm nx agent-dev story

yarn nx agent-dev story

npx nx agent-dev story

bunx nx agent-dev story

Then, in a second terminal, start the chat:

pnpm nx agent-chat story

yarn nx agent-chat story

npx nx agent-chat story

bunx nx agent-chat story

Your first message should tell the agent your hero’s name and the genre (for example: My name is Alice. Start my zombie adventure.) and the story will stream back.

Congratulations. You have built and tested your first Strands Agent locally! 🎉🎉🎉