Strands AgentCore Tools: Self-Deploying Agents ☁️

Agentic tools that enable AI agents to autonomously deploy, manage, and monitor themselves on AWS Bedrock AgentCore.

9 Python functions that wrap AWS Bedrock AgentCore boto3 APIs for deployment, invocation, monitoring, and lifecycle management.

Installation

pip install strands-agentcore-tools

What This Is

Tools for: - ✅ Deploy agents to AgentCore from code - ✅ Invoke deployed agents programmatically - ✅ Monitor CloudWatch logs and metrics - ✅ Manage memory, sessions, and OAuth - ✅ Integrate with Strands agents for autonomous deployment

Quick Start

1. Create Your Agent

# agent.py
import threading
from bedrock_agentcore import BedrockAgentCoreApp
from strands import Agent

app = BedrockAgentCoreApp()

@app.entrypoint
async def invoke(payload, context):
    """Async entrypoint with streaming support"""
    agent = Agent()
    prompt = payload.get("prompt", "")
    mode = payload.get("mode", "streaming")

    if mode == "fire_and_forget":
        task_id = app.add_async_task("agent_processing", payload)
        thread = threading.Thread(
            target=run_agent_background,
            args=(agent, prompt, task_id),
            daemon=True
        )
        thread.start()
        yield {"status": "started", "content": [{"text": "Running in background"}]}

    elif mode == "sync":
        result = agent(prompt)
        yield {"status": "success", "content": [{"text": str(result)}]}

    else:  # streaming
        for event in agent.stream_async(prompt):
            yield event

def run_agent_background(agent, prompt, task_id):
    """Background worker"""
    try:
        for event in agent.stream_async(prompt):
            print(event)
        app.complete_async_task(task_id)
    except Exception as e:
        app.logger.error(f"Background task failed: {e}")
        app.complete_async_task(task_id)

app.run()

2. Deploy It

from strands_agentcore_tools import configure, launch, invoke

# Configure
configure(action="configure", entrypoint="agent.py", agent_name="my-agent")

# Launch
launch(action="launch", agent_name="my-agent")

# Invoke
invoke(
    agent_arn="arn:aws:bedrock-agentcore:...",
    payload='{"prompt": "Hello!", "mode": "streaming"}'
)

Available Tools

Tool	Purpose
configure	Configure agent for deployment
launch	Deploy to AgentCore
invoke	Call deployed agent
agents	List/manage agents
status	Check agent status
logs	View CloudWatch logs
memory	Manage agent memory
identity	Configure OAuth
session	Manage sessions

Architecture

flowchart TB
    subgraph Local["💻 Local Development"]
        Code[agent.py]
        Tools[AgentCore Tools]
    end

    subgraph AWS["☁️ AWS"]
        CB[CodeBuild]
        ECR[ECR]
        AC[AgentCore Runtime]
        CW[CloudWatch]
    end

    Code --> Tools
    Tools -->|configure| CB
    CB -->|build| ECR
    ECR -->|deploy| AC
    AC -->|logs| CW

    style AC fill:#f9f,stroke:#333,stroke-width:4px

Agent with Memory

# agent.py - Agent with STM + LTM
from bedrock_agentcore import BedrockAgentCoreApp
from strands import Agent
from bedrock_agentcore.memory.integrations.strands.config import (
    AgentCoreMemoryConfig, RetrievalConfig
)
from bedrock_agentcore.memory.integrations.strands.session_manager import (
    AgentCoreMemorySessionManager
)

app = BedrockAgentCoreApp()
MEMORY_ID = os.getenv("BEDROCK_AGENTCORE_MEMORY_ID")

@app.entrypoint
async def invoke(payload, context):
    session_id = context.session_id
    actor_id = context.headers.get("X-Amzn-Bedrock-AgentCore-Runtime-Custom-Actor-Id", "user")

    # Configure memory with semantic search
    memory_config = AgentCoreMemoryConfig(
        memory_id=MEMORY_ID,
        session_id=session_id,
        actor_id=actor_id,
        retrieval_config={
            f"/users/{actor_id}/facts": RetrievalConfig(top_k=3, relevance_score=0.5)
        }
    )

    agent = Agent(
        session_manager=AgentCoreMemorySessionManager(memory_config)
    )

    async for event in agent.stream_async(payload.get("prompt", "")):
        yield event

CLI Usage

# Configure
strands-agentcore configure --entrypoint agent.py --agent-name my-agent

# Launch
strands-agentcore launch --agent-name my-agent

# Check status
strands-agentcore status --agent-name my-agent

# View logs
strands-agentcore logs --agent-name my-agent --tail

# Invoke
strands-agentcore invoke --agent-arn arn:... --prompt "Hello!"

---

Zero DevOps deployment—agents deploy themselves. ☁️