aws-integration-testing

AWS Integration Testing Skill

Generate integration tests by reading handler code, identifying every AWS service interaction, and producing tests with correct moto mocks for each one.

How to Generate Integration Tests

Phase 1: Analyze AWS Interactions

Read the handler source code and find every AWS interaction:

# Look for these patterns in the handler:

# Direct client creation
client = boto3.client("s3")
resource = boto3.resource("dynamodb")
sqs = boto3.client("sqs")
sns = boto3.client("sns")
events = boto3.client("events")
kinesis = boto3.client("kinesis")

# Method calls on clients
client.get_object(Bucket=..., Key=...)
client.put_object(Bucket=..., Key=..., Body=...)
table.get_item(Key=...)
table.put_item(Item=...)
table.query(KeyConditionExpression=...)
sqs.send_message(QueueUrl=..., MessageBody=...)
sqs.send_message_batch(QueueUrl=..., Entries=...)
sns.publish(TopicArn=..., Message=...)
events.put_events(Entries=[...])
kinesis.put_record(StreamName=..., Data=..., PartitionKey=...)

# Error handling around AWS calls
try:
    client.get_object(...)
except ClientError as e:
    if e.response["Error"]["Code"] == "NoSuchKey":
        ...

Build a table of interactions:

Service	Method	Parameters	Success Path	Error Path
s3	get_object	Bucket, Key	Returns Body	NoSuchKey → 404
s3	put_object	Bucket, Key, Body	Returns None	AccessDenied → 403
dynamodb	get_item	Key	Returns Item	Item not found → None
sqs	receive_message	QueueUrl	Returns Messages	Timeout → empty
sns	publish	TopicArn, Message	Returns MessageId	InvalidParameter → 400
events	put_events	Entries	FailedEntryCount=0	InternalFailure → retry
kinesis	put_record	StreamName, Data, PartitionKey	Returns ShardId	ProvisionedThroughputExceeded → 429

Phase 2: Generate moto-based Tests

For each AWS service interaction, generate test(s). Use the patterns in:

• references/lambda-s3-patterns.md
• references/lambda-dynamodb-patterns.md
• references/lambda-rds-patterns.md
• references/lambda-sqs-patterns.md
• references/lambda-sns-patterns.md
• references/lambda-eventbridge-patterns.md
• references/lambda-kinesis-patterns.md

General Structure

"""Integration tests for {handler_name}. Auto-generated from code analysis."""
import json
import boto3
import pytest
from moto import mock_aws
from unittest.mock import patch, MagicMock

# Discovered from handler's os.environ/os.getenv calls:
ENV_VARS = {
    "BUCKET_NAME": "test-bucket",
    "TABLE_NAME": "test-table",
    "AWS_DEFAULT_REGION": "us-east-1",
}

REGION = "us-east-1"


@pytest.fixture(autouse=True)
def set_env(monkeypatch):
    for k, v in ENV_VARS.items():
        monkeypatch.setenv(k, v)


@pytest.fixture
def aws_credentials(monkeypatch):
    monkeypatch.setenv("AWS_ACCESS_KEY_ID", "testing")
    monkeypatch.setenv("AWS_SECRET_ACCESS_KEY", "testing")
    monkeypatch.setenv("AWS_SECURITY_TOKEN", "testing")
    monkeypatch.setenv("AWS_DEFAULT_REGION", REGION)


@pytest.fixture
def mock_context():
    ctx = MagicMock()
    ctx.function_name = "test-function"
    ctx.memory_limit_in_mb = 256
    ctx.get_remaining_time_in_millis.return_value = 30000
    return ctx

Phase 3: Test Every Code Path

For each code path discovered in Phase 1, generate one test:

# Happy path — AWS call succeeds
def test_handler_reads_s3_and_returns_data(aws_credentials, mock_context):
    with mock_aws():
        # ARRANGE: Create the AWS resource with test data
        s3 = boto3.client("s3", region_name=REGION)
        s3.create_bucket(
            Bucket="test-bucket",
            CreateBucketConfiguration={"LocationConstraint": REGION},
        )
        s3.put_object(Bucket="test-bucket", Key="data.json", Body=b'{"key": "value"}')

        # ACT: Call the handler
        from handler_module.main import lambda_handler
        event = {"pathParameters": {"key": "data.json"}}
        result = lambda_handler(event, mock_context)

        # ASSERT: Verify handler used S3 data correctly
        assert result["statusCode"] == 200
        body = json.loads(result["body"])
        assert body["key"] == "value"

# Error path — S3 object not found
def test_handler_missing_s3_key_returns_404(aws_credentials, mock_context):
    with mock_aws():
        s3 = boto3.client("s3", region_name=REGION)
        s3.create_bucket(
            Bucket="test-bucket",
            CreateBucketConfiguration={"LocationConstraint": REGION},
        )
        # Don't create the object — it's missing

        from handler_module.main import lambda_handler
        event = {"pathParameters": {"key": "missing.json"}}
        result = lambda_handler(event, mock_context)

        assert result["statusCode"] == 404

# Error path — unexpected exception
def test_handler_unexpected_error_returns_500(mock_context):
    with patch("handler_module.main.boto3") as mock_boto:
        mock_boto.client.return_value.get_object.side_effect = Exception("boom")

        from handler_module.main import lambda_handler
        result = lambda_handler({"pathParameters": {"key": "x"}}, mock_context)

        assert result["statusCode"] == 500

Key Principles

1. One test per code path — if the handler has 5 branches, write 5+ tests
2. Real moto mocks, not unittest.mock — use mock_aws() for AWS services so real API semantics are tested
3. Set up real state — if the handler reads from S3, put an actual object in moto S3
4. Test error paths explicitly — NoSuchKey, ConditionalCheckFailed, throttling
5. Import the handler inside the test — avoid module-level import caching issues with moto
6. Monkeypatch env vars — match exactly what the handler reads with os.getenv()

Markers

pytestmark = pytest.mark.integration

Commands

# All integration tests
pytest tests/integration/ -v -m integration

# Moto-only (no Docker needed)
pytest tests/integration/ -v -m integration -k "not testcontainers"

# With testcontainers (needs Docker)
pytest tests/integration/ -v -m integration