How to Build an AI Agent with AWS Bedrock

Install the Bedrock AgentCore Python package, the Strands Agents SDK, and the agentcore CLI, which is distributed through npm rather than pip since it's a standalone deployment tool, not a Python library. Confirm you have at least one Claude model enabled under Bedrock's Model access page in your account's region before continuing; AgentCore Runtime won't help if the underlying model call is denied.

1pip install bedrock-agentcore strands-agents boto3
2npm install -g @aws/agentcore
3aws configure   # set your AWS access key, secret, and region

Write get_shipment_status with Strands' @tool decorator. Its docstring is what the model reads to decide when to call it, so describe exactly what it returns. This tool only reads from your shipments API; keep any write logic entirely out of it, the same separation of concerns that makes the next tool's approval gate meaningful.

1from strands import Agent, tool
2import httpx
3
4@tool
5def get_shipment_status(shipment_id: str) -> dict:
6    """Look up a shipment's current status, carrier, and expected delivery date."""
7    resp = httpx.get(f"https://api.example-store.com/shipments/{shipment_id}")
8    resp.raise_for_status()
9    return resp.json()

issue_shipping_credit takes an approved flag that defaults to False. When it's False, the tool returns a drafted amount and reason without charging anything; when a human has reviewed the draft and the next invoke passes approved=True, it actually calls your payments API. The function itself enforces the gate, so no amount of clever prompting in the conversation can skip it.

1@tool
2def issue_shipping_credit(shipment_id: str, amount: float, reason: str, approved: bool = False) -> dict:
3    """Draft or issue a shipping credit. Only issues the credit when approved is True;
4    otherwise returns a draft for a human to review."""
5    if not approved:
6        return {"status": "drafted", "amount": amount, "reason": reason,
7                "note": "A support lead must confirm this before it's issued."}
8    # In production this calls your payments API (e.g. Stripe credit notes)
9    return {"status": "issued", "amount": amount, "reason": reason}

Wrap both tools in a Strands Agent with a system prompt that states the rule in plain language: always check status before promising anything, and never tell the customer a credit is confirmed unless the tool actually returned status issued. Strands talks to Bedrock directly; leave model unset to use your account's available Claude model, or pass model with the exact Bedrock model ID shown on your Model access page if you need to pin one.

1SYSTEM_PROMPT = (
2    "You are a shipment support agent. Always call get_shipment_status before "
3    "discussing a shipment. If a delay justifies a credit, call issue_shipping_credit "
4    "with approved left as False and tell the customer a support lead will confirm it. "
5    "Never say a credit was issued unless the tool result says status: issued."
6)
7
8agent = Agent(
9    system_prompt=SYSTEM_PROMPT,
10    tools=[get_shipment_status, issue_shipping_credit],
11)

BedrockAgentCoreApp is the only AWS-specific code in this guide. Its @app.entrypoint decorator marks the function AgentCore Runtime calls on every invoke; read the prompt and the approved flag out of the incoming request and pass them through to the Strands agent. app.run() starts the server AgentCore's Runtime expects.

1from bedrock_agentcore import BedrockAgentCoreApp
2
3app = BedrockAgentCoreApp()
4
5@app.entrypoint
6def handler(request):
7    prompt = request.get("prompt", "")
8    approved = request.get("approved", False)
9    if approved:
10        prompt += " The pending credit has been approved by a support lead; issue it now."
11    result = agent(prompt)
12    return {"result": str(result)}
13
14app.run()

agentcore create scaffolds the project configuration around your existing file, agentcore build packages it, and agentcore deploy ships it to AgentCore Runtime. Use agentcore dev first if you want to test against a browser-based Agent Inspector before deploying for real.

1agentcore create --name shipment-support --template agent-python-strands
2agentcore dev --mode browser   # optional local test with Agent Inspector
3agentcore build
4agentcore deploy

Call the deployed agent with approved left out (or false) and confirm you get a drafted credit, not an issued one. Then simulate the human review step, a support lead reading the draft in your own ticketing tool, and send a second invoke with approved: true for the same shipment and amount. Only this second call should show status: issued in the tool's output.

1agentcore invoke --runtime shipment-support --payload '{"prompt": "SHIP-204 is 3 days late, can I get a credit?"}'
2# ... support lead reviews the drafted amount ...
3agentcore invoke --runtime shipment-support --payload '{"prompt": "Issue the credit for SHIP-204", "approved": true}'