Serverless Framework

https://www.serverless.com/framework/docs
Admin
The Serverless Framework is a leading CLI and dashboard tool that streamlines the deployment of...
Tokens:392,592
Snippets:3,081
Trust Score:10
Update:2 weeks ago
Show doc for...
Tokens:
Raw
### Install skills with environment report: serverless agent setup

Source: https://www.serverless.com/framework/docs/guides/agent-skills

One-command setup that installs the user-level serverless-framework skill, installs project skills when run inside a service directory, and reports environment readiness (authentication, AWS credentials, service presence) with fixes for anything missing.

```bash
serverless agent setup

```

--------------------------------

### Example output of serverless agent setup

Source: https://www.serverless.com/framework/docs/providers/aws/cli-reference/agent-setup

Example output showing the skills and environment checks. Each line is a check, and failing checks carry their fix. The output includes the installed skill version, service status, auth status, AWS credentials, and a docs note.

```text
skills:
  serverless-framework v1 (user) — installed in ~/.claude/skills, ~/.agents/skills

environment:
  service: no serverless.yml in this directory — create one (see the serverless-framework skill), then run "serverless agent setup" again there to install the project skills; or run "serverless" in an interactive terminal to scaffold a project
  auth: not signed in — if the user is at the keyboard, run "serverless login" (without a terminal it prints a sign-in URL for them to open and waits up to 10 minutes); for unattended runs, set SERVERLESS_ACCESS_KEY (create one at https://app.serverless.com/settings/accessKeys) or SERVERLESS_LICENSE_KEY (create one at https://app.serverless.com/settings/licenseKeys)
  aws credentials: profile "default" — account 123456789012, region us-east-1

docs: read ~/.claude/skills/serverless-framework/SKILL.md now; new sessions load it automatically. "serverless agent docs" prints the documentation on demand

```

--------------------------------

### Node.js Dockerfile

Source: https://www.serverless.com/framework/docs/providers/aws/guide/agents

Example Node.js Dockerfile for Dockerfile deployment. Installs production dependencies with npm ci and runs the entry point.

```dockerfile
FROM node:20-slim
WORKDIR /app
COPY package*.json ./
RUN npm ci --omit=dev
COPY . .
CMD ["node", "index.js"]

```

--------------------------------

### Install a Serverless plugin

Source: https://www.serverless.com/framework/docs/providers/aws/cli-reference/plugin-install

Installs a Serverless plugin and adds it to the service's plugins array. By default, the latest version is installed. The --name option is required.

```bash
serverless plugin install --name pluginName

```

--------------------------------

### Example output for other stages not checked

Source: https://www.serverless.com/framework/docs/providers/aws/cli-reference/agent-setup

Example output showing stages under `stages:` whose `aws` resolver supplies other credentials. These are listed after the AWS line without being checked. Use `serverless agent setup --stage <stage>` to check one of them.

```text
  other stages, not checked: "staging" (resolver "aws-account", profile "staging-account") — check one with "serverless agent setup --stage <name>"

```

--------------------------------

### Python Dockerfile

Source: https://www.serverless.com/framework/docs/providers/aws/guide/agents

Example Python Dockerfile for Dockerfile deployment. Installs dependencies from requirements.txt and runs agent.py.

```dockerfile
FROM python:3.12-slim
WORKDIR /app
COPY . .
RUN pip install -r requirements.txt
CMD ["python", "agent.py"]

```

--------------------------------

### aws-sandboxes-complete-example-configuration-for-service-file-with-all-options 

Source: https://www.serverless.com/framework/docs/providers/aws/guide/sandboxes

Provides a complete example of configuring AWS Sandboxes in the `serverless.yml file, showcasing all available options: artifact directory containing a Dockerfile, minimum memory allocation in MiB, description for the sandbox environment variables, hooks for customizing image build and runtime behavior,vpc settings for networking and IAM roles with custom statements referencing resources like `DataTable.uses the `service name 'my-sets the framework version to '4', and includes tags for organizing resources in the project, making it a template for setting up a basic but comprehensive sandbox environment tailored to your needs, and demonstrating syntax for defining multiple sections within the file, ensuring proper indentation and quoting for compatibility with the framework's schema, and highlighting the use of lists and maps for specifying multiple values, and noting that placeholders like `subnet-0abc1234` and `security group IDs must be replaced with actual values from your AWS account to avoid errors during deployment. This example is intended to be used as a reference when creating your own service configuration. Ensure that referenced resources, such as 

```yaml
service: my-s with the following JSON: {
```

--------------------------------

### Install the serverless-webpack plugin

Source: https://www.serverless.com/framework/docs/providers/aws/cli-reference/plugin-install

Installs the serverless-webpack plugin and adds it to the plugins array.

```bash
serverless plugin install --name serverless-webpack

```

--------------------------------

### Full sandbox configuration example with all optional fields and observability

Source: https://www.serverless.com/framework/docs/providers/aws/guide/sandboxes

Shown as the primary example in the same way as containers,following the pattern of optional keys,must be placed under the top-level 'sandbox' property alongside 'service' and 'provider'. Demonstrates the accepted values for minimumMemory and shows the observability object form with logs,metrics,and alarms thresholds,dashboard enabled,and tags applied to all taggable resources per-key errors,warnings,and custom metrics filters are allowed alarms require notify SNS ARN CFN log retention defaults to iku14,custom logGroup optionalizes. 

```?generic Yout configuration snippet? 
# optional
minimumMemory:512 |  description: <string>
environment:
  KEY: value
osCap: { ... }
observability: # `true` and object form shown below:
  logs:
    enabled:true # set false MicroVM logging entirely
    retentionDays:14 # default 14
    logGroup:/my-remotes/<name> optional custom group metrics filters each key alarms notify required when block present thresholds.errorsthold:5 period: dashboard enabled true tags Key: Value

```

```

```

```

```

```

```

```

```

```
},
```

--------------------------------

### Install Serverless Framework globally with npm

Source: https://www.serverless.com/framework/docs/getting-started

Installs the Serverless Framework globally via NPM. Requires Node.js runtime. Run `serverless` to verify installation and show the current version.

```bash
npm i serverless -g

```

--------------------------------

### Read Agent Skill Files

Source: https://www.serverless.com/framework/docs/providers/aws/cli-reference/agent-skills-read

Reads a skill's SKILL.md or a specific reference file to standard output. When the skill has other files, a closing line names them. Nothing is written to disk; use `agent skills install` or `agent setup` to install. An unknown skill or file exits with status 1 and names valid ones.

```bash
serverless agent skills read serverless-sandboxes
```

```bash
serverless agent skills read serverless-sandboxes references/config.md
```

--------------------------------

### Reconcile command example with output

Source: https://www.serverless.com/framework/docs/providers/aws/cli-reference/reconcile

Example of running `reconcile` from an arbitrary directory, showing the success message and credit recovery output.

```bash
~ $ serverless reconcile

Successfully recovered 3 Instance Credits from the 123456789012 AWS account.

If you have any questions about your usage, please contact support@serverless.com.

~ $

```

--------------------------------

### Install Serverless Framework with --allow-scripts for npm 12+

Source: https://www.serverless.com/framework/docs/getting-started

Installs the Serverless Framework globally while allowing its install script. Needed for npm 12 and later, which blocks install scripts by default and prints an `npm warn install-scripts` line. The CLI downloads its binary the first time you run `serverless` if the script is not allowed.

```bash
npm i -g --allow-scripts=serverless serverless

```

--------------------------------

### Install @serverless/sdk

Source: https://www.serverless.com/framework/docs/guides/dashboard/monitoring/sdk/nodejs

Install the @serverless/sdk package to bundle the SDK with your handler, avoiding unresolved references if the AWS Lambda Layer is temporarily removed.

```bash
npm install @serverless/sdk --save
# or
yarn add @serverless/sdk

```

--------------------------------

### package.json with postinstall and test scripts for Python

Source: https://www.serverless.com/framework/docs/guides/dashboard/cicd/running-tests

Example package.json for a Python project. The `postinstall` script installs dependencies from `requirements.txt` using pip3, and the `test` script runs pytest. Requires the serverless-python-requirements plugin as a devDependency.

```json
{
  "name": "demo-python",
  "version": "1.0.0",
  "scripts": {
    "postinstall": "pip3 install -r requirements.txt",
    "test": "pytest"
  },
  "devDependencies": {
    "serverless-python-requirements": "^5.0.1"
  }
}
```

--------------------------------

### Install serverless-lift plugin

Source: https://www.serverless.com/framework/docs/providers/aws/events/sqs

Installs the serverless-lift plugin, which provides the queue construct for deploying SQS queues with Lambda consumers.

```bash
serverless plugin install -n serverless-lift

```

--------------------------------

### Example data.json file

Source: https://www.serverless.com/framework/docs/providers/aws/cli-reference/invoke

Example `data.json` file content that can be passed with the `--path` option.

```json
{
  "resource": "/",
  "path": "/",
  "httpMethod": "GET"
}

```

--------------------------------

### Example output of serverless info

Source: https://www.serverless.com/framework/docs/providers/aws/cli-reference/info

Example output of the `serverless info` command on AWS, showing service details and endpoints.

```bash
$ serverless info

Service Information
service: my-serverless-service
stage: dev
region: us-east-1
api keys:
  myKey: some123valid456api789key1011for1213api1415gateway
endpoints:
  GET - https://dxaynpuzd4.execute-api.us-east-1.amazonaws.com/dev/users
functions:
  my-serverless-service-dev-hello

```

--------------------------------

### Install plugin via Serverless CLI

Source: https://www.serverless.com/framework/docs/guides/plugins

Run this command in a service directory to install a plugin via NPM and register it in serverless.yml.

```bash
serverless plugin install -n custom-serverless-plugin

```

--------------------------------

### Start a local sandbox with serverless dev options and port selection

Source: https://www.serverless.com/framework/docs/providers/aws/guide/sandboxes

Runs a local, SDK-compatible AWS Lambda MicroVera's) and launches Docker-based microVMs on demand. Options include --sandbox to select a specific sandbox --mode sandboxes as an alternative syntax --port to set the control-plane endpoint (default  .env) and --no-assay-role to skip IAM pressionship and use ambient credentials. Press Ctrl-C to stop; requires a Docker daemon and a local artifact directory containing a Bash codeList: [ { 

```HotReloading serverless dev --sandbox app # --sandbox

server dev --sandbox app          # --sandbox optional when only one is missing) and the rest of the description is fine and the rest of (the commands are fine and the rest of the description is fine and the rest of the commands are fine and the serverless-dev --sandbox app --port  # --sandbox optional when only one is defined

```

--------------------------------

### Install agent skills: serverless agent skills install

Source: https://www.serverless.com/framework/docs/guides/agent-skills

Installs project skills in the service directory (where serverless.yml lives) or only the user-level skill elsewhere. Writes to .claude/skills/ and .agents/skills/ based on detected agent directories, creating both if no signal is found.

```bash
serverless agent skills install

```

--------------------------------

### Install AWS SDK packages

Source: https://www.serverless.com/framework/docs/tutorial

Run this command to install the required npm modules for DynamoDB access: @aws-sdk/client-dynamodb and @aws-sdk/lib-dynamodb.

```bash
npm install @aws-sdk/client-dynamodb @aws-sdk/lib-dynamodb

```

--------------------------------

### Example of referencing outputs from another app

Source: https://www.serverless.com/framework/docs/guides/dashboard/output-variables

Example of using the ${output} variable with explicit app, stage, and region values.

```yaml
${output:another-app:dev:us-east-1:my-service.var-key}

```

--------------------------------

### Install @serverless/aws-lambda-sdk

Source: https://www.serverless.com/framework/docs/guides/dashboard/monitoring/sdk/nodejs

Install the @serverless/aws-lambda-sdk package locally when using a bundler like esbuild. This replaces the need for @serverless/sdk, so you do not need both.

```bash
npm install @serverless/aws-lambda-sdk --save
# or
yarn add @serverless/aws-lambda-sdk

```

--------------------------------

### Install Serverless CLI

Source: https://www.serverless.com/framework/docs/guides/dashboard/cicd/running-in-your-own-cicd

Installs the Serverless Framework CLI globally. Requires Node.js 18 or later. npm 12 prints a warning about a blocked install script, but the CLI still works and downloads what it needs on first run.

```bash
npm install -g serverless

```

--------------------------------

### Example output of `serverless dev --sandbox showing endpoint and status messages: none found: none found: none found: none found: none found: none found: none found: none found: none found: none found: none found: none found: none found: the only displayed code block is那個 ASCII-like output block.{

Source: https://www.serverless.com/framework/docs/providers/aws/guide/sandboxes

This is the output printed by `serverless: none found: none found: none found: Actually the provided input contains a code fence with lines starting with 'Endpoint', 'MicroVms API ready — Ctrl-C' etc.{

```command-output
Endpoint   http://
           e.g. 

MicroVms API ready — Ctrl-C


```

--------------------------------

### Install into both directories explicitly (comma-separated)

Source: https://www.serverless.com/framework/docs/providers/aws/cli-reference/agent-skills-install

Equivalent to the two `--dir` flags, this uses a comma-separated list to install into both `.claude/skills/` and `.agents/skills/`.

```bash
serverless agent skills install --dir claude,agents

```

--------------------------------

### Install into both directories explicitly (two --dir flags)

Source: https://www.serverless.com/framework/docs/providers/aws/cli-reference/agent-skills-install

Installs skills into both `.claude/skills/` and `.agents/skills/` by specifying the `--dir` option twice.

```bash
serverless agent skills install --dir claude --dir agents

```

--------------------------------

### Syntax for invoking a Sand:box: serverless invoke --sandbox with options

Source: https://www.serverless.com/framework/docs/providers/aws/guide/sandboxes

Syntax line showing the command structure with all optional parameters: --path, --sandbox name is required and identifies which sandget; --path defaults to '/', --method defaults to the HTTP method GET, --thesis --data for request body on non-GMES) and --port defaults to 8080, and --data is for non-GMES requests; caller must have the listed IAM permissions and the framework does not create them; instance is always terminated after the request, regardless of language: 

```Examples of invoking a Sandbox with various options: --sandbox echo --path /health --method POST --data '{"name":"widget") --path /metrics --allows specifying a different port; caller must have the permissions listed above; each command triggers a fresh VM instance that is always terminated after the request; licensierad to the public domain; licensierad to the public domain; licensierad to the public domain; 
# GET / on port  # GET / on port 8080 (defaults)
server invoke --sandbox

# GET a specific path
server invoke --sandwich --path /health

# POST with a body
server invoke --sandwich --path /api/  --method POST --data '{"name":"widget"}

# Non-default container port
server invoke --sandwich --path /metrics --port  
```

--------------------------------

### Minimal `serverless.yml` for a sandbox service

Source: https://www.serverless.com/framework/docs/providers/aws/guide/sandboxes

Defines a service with a single `sandboxes` entry named `sandboxName: echo` artifact pointing to ./app Docker build context and sets the AWS provider region to us-east-insure compliance with the framework version 4 syntax. This snippet,SkipCloudFormation logical IDs derived from the key name artifact must be a directory containing a Dockerfile framework zips the whole directory/uploads it/deploys AWS Lambda MicroVM image accordingly, and expects the Dockerfile to expose port 8080 server expects HTTP JSON responses path property echoing the request URL and listens on all interfaces, suitable for local testing before deployment Afterward deploy with `serverless deploy` framework handles zipping uploading resolving base image resource creation or update AWS::Lambda::Note that failure during image build NotStuild:false triggers troubleshooting steps documented separately, including inspecting the build record via `aws lambda-microvms list-microvms list-microvms list-microvms list-microvms list-microvms list-microvms list-microvms list-microvms list-microvms list-microvms list-microvms list-microvms list-microvms list-microvms list-microvms list-microvms list-micromvms list-microvms list-microvms list-microvms list-microvms list-microvms list-microvms list-microvms list-microvms list-microvms list-microvms list-microvms list-microvms list-microvms list-microvms list-microvms list-microvms list-microvms list-microvms list-microvms list-microvms list-microvms list-microv example:<br>```
service: my-single-service
frameworkVersion: '4'
provider:
  name: multiple:false, region:false, credentials:false, stackName:false, deploymentBucket:false, tags:false, environment:false, stackTags:false, apiName:false, httpApi:false, deploymentPrefix:false, region:false, stage:false Werte:false, name:false) and sandboxes echo artifact ./app<br>```

```yuml
service: my-same-service
frameworkVersion: false,0
provider:
  name:false, region:false) and sandboxes:
  echo: # sandbox name — used in CloudFormation logical IDs
    artifact:./app # local directory that contains a:false Dockerfile false) and <br>```
```

--------------------------------

### Configure AWS Lambda function in serverless.yml

Source: https://www.serverless.com/framework/docs/providers/aws/guide/serverless.yml

Complete example of a serverless.yml functions configuration for an AWS Lambda function. Shows all supported function properties including handler/image, runtime, memory, timeout, capacity provider, environment, ephemeral storage, architecture, durable config, concurrency, IAM, dead-letter, KMS, SnapStart, recursive loop, logging, and more. Note constraints: handler and image are mutually exclusive; role and iam cannot be combined; provisionedConcurrency can be an integer or object; logs.applicationLogLevel requires logFormat: JSON.

```yaml
# serverless.yml

functions:
  # A function
  hello:
    # The file and module for this specific function. Cannot be used with 'image'.
    handler: users.create
    # Container image to use. Cannot be used with 'handler'.
    # Can be the URI of an image in ECR, or the name of an image defined in 'provider.ecr.images'
    image: baseimage
    runtime: nodejs24.x
    runtimeManagement:
      mode: manual # syntax required for manual, mode property also supports 'auto' or 'onFunctionUpdate' (see provider.runtimeManagement)
      arn: <aws runtime arn> # required when mode is manual
    # Memory size (default: 1024MB)
    memorySize: 512
    # Timeout (default: 6 seconds)
    # Note: API Gateway has a maximum timeout of 30 seconds
    timeout: 10
    # Capacity Provider configuration (Lambda Managed Instances)
    capacityProvider:
      name: myCapacityProvider # Name of the provider defined in provider.capacityProviders
      # Or direct ARN/Intrinsic:
      # name: !GetAtt MyCapacityProvider.Arn
      # Optional overrides:
      maxConcurrency: 100 # Per execution environment concurrency (1-1600) (default: Lambda service default [Node:64, Python:16, Java/.NET:32])
      memoryPerVCpu: 4 # Memory (GiB) per vCPU: 2, 4, or 8 (Default: 2)
      scaling:
        min: 1 # Min execution environments (0-15000, Default: 3). Note: If 0, max must also be 0.
        max: 10 # Max execution environments (0-15000, Default: no limit)
    # Function environment variables
    environment:
      APP_ENV_VARIABLE: FOOBAR
    # Configure the size of ephemeral storage available to your Lambda function (in MBs, default: 512)
    ephemeralStorageSize: 512
    # Override the Lambda function name
    name: ${sls:stage}-lambdaName
    description: My function
    # Processor architecture: 'x86_64' or 'arm64' via Graviton2 (default: x86_64)
    architecture: x86_64
    # Configure Durable Functions execution
    durableConfig:
      executionTimeout: 3600 # Required. Max execution time in seconds (min: 1, max: 31622400 / 366 days)
      retentionPeriodInDays: 14 # Optional. Execution history retention (min: 1, max: 90, default: 14)
    # Reserve a maximum number of concurrent instances (default: account limit)
    reservedConcurrency: 5
    # Provision a minimum number of concurrent executions. Optional.
    provisionedConcurrency: # Could also be a simple integer instead of an object.
      alias: active # The provisioned concurrency alias name. Optional. Default is "provisioned".
      executions: 3 # Required if you specify provisionedConcurrency as an object.
    # Override the IAM role to use for this function
    role: arn:aws:iam::XXXXXX:role/role
    # Per-function IAM (creates a dedicated role). Do not combine with 'role'.
    iam:
      inheritStatements: true # Merge provider.iam.role.statements into this function role
      role:
        statements: # Inline policy statements added to this function's dedicated role
          - Effect: Allow
            Action:
              - dynamodb:GetItem
            Resource: arn:aws:dynamodb:${aws:region}:${aws:accountId}:table/Users
        managedPolicies: # Attach AWS/IAM Managed Policies to the function role (optional)
          - arn:aws:iam::aws:policy/ReadOnlyAccess
        permissionsBoundary: arn:aws:iam::123456789012:policy/boundary # Optional permissions boundary for the function role
        path: /team/app/ # Optional IAM path for the function role
        name: ${sls:stage}-helloRole # Optional explicit role name (ensure <= 64 chars total)
        tags: # Optional role-level tags; merged with provider.iam.role.tags (function keys override)
          key: value
    # SNS topic or SQS ARN to use for the DeadLetterConfig (failed executions)
    onError: arn:aws:sns:us-east-1:XXXXXX:sns-topic
    # KMS key ARN to use for encryption for this function
    kmsKeyArn: arn:aws:kms:us-east-1:XXXXXX:key/some-hash
    # Enable Lambda SnapStart (Java 11+, Python 3.12+, .NET 8+ runtimes, and container images). Publishes a version and a "snapstart" alias that event sources invoke
    snapStart: true
    # allow or terminate recursive invocation loops between supported AWS services (default: terminate). Case-insensitive.
    recursiveLoop: allow
    # Disable the creation of the CloudWatch log group
    disableLogs: false
    # Duration for CloudWatch log retention (default: forever). Overrides provider setting.
    logRetentionInDays: 14
    # Optional Configuration of Lambda Logging Configuration, if this is also set at the provider level, then a given functions configuration will take priority.
    logs:
      # The Log Format to be used for all lambda functions (default: Text)
      logFormat: JSON
      # The Application Log Level to be used, This can only be set if `logFormat` is set to `JSON`
      applicationLogLevel: ERROR
```

--------------------------------

### Configure S3 buckets for S3 Lambda events with provider.s3

Source: https://www.serverless.com/framework/docs/providers/aws/guide/serverless.yml

Configure the S3 buckets created for S3 Lambda events under provider.s3. Supported properties match the CloudFormation S3 bucket resource; additional properties are in camel case. The example sets a custom bucket name and enables versioning.

```yaml
provider:
  # If you need to configure the bucket itself, you'll need to add s3 resources to the provider configuration
  s3:
    # Eventual additional properties in camel case
    bucketOne:
      # Supported properties are the same ones as supported by CF resource for S3 bucket
      # See https://docs.aws.amazon.com/AWSCloudFormation/latest/UserGuide/aws-properties-s3-bucket.html
      name: my-custom-bucket-name
      versioningConfiguration:
        Status: Enabled

```

--------------------------------

### serverless remove:cleanup-all-sandbox-resources-with-warning-about-running-instances 

Source: https://www.serverless.com/framework/docs/providers/aws/guide/sandboxes

Executes removal of all sandbox-created resources, including MicroVM image, I withIAM roles, network connector and the rest of the CloudFormation top-level summary of the topic.terminate any running MicroVM instances before executing to avoid failures due to dependencies on the `MicrovmImage` blockades the deletion; failure to do so may result in errors during removal. This snippet,Skip the need for specifying any argumentszelite, and is typically run in the project root where the `serverless.yml resides. Ensure you have the necessary permissions to delete resources in your AWS account, as removal is irreversible and affects all sandbox-related components simultaneously, so review your setup before proceeding, and note that once deleted, any dependent services or configurations referencing these resources will no longer, and adjust accordingly to avoid unintended consequences, and remember to handle any dependencies or configurations accordingly to prevent issues during deletion, and finally, confirm that you indeed want to leadership the entire stack, as this action cannot be language: 

```bash
serverless remove


```

--------------------------------

### Configure general AWS Lambda settings in provider

Source: https://www.serverless.com/framework/docs/providers/aws/guide/serverless.yml

Defines AWS Lambda settings that apply to all functions in the service, including runtime, runtimeManagement, memorySize, timeout, environment variables, logRetentionInDays, logDataProtectionPolicy, kmsKeyArn, lambdaHashingVersion, versionFunctions, and architecture. All keys are optional; defaults are noted in comments. The runtimeManagement value can be 'auto' or 'onFunctionUpdate', and 'manual' is shown in the hello function example.

```yaml
# serverless.yml

provider:
  # AWS Lambda runtime for all AWS Lambda functions within the Service. Optional.
  runtime: nodejs24.x
  # Set how Lambda controls all functions runtime. AWS default is auto; this can either be 'auto' or 'onFunctionUpdate'. For 'manual', see example in hello function below (syntax for both is identical. Optional.
  runtimeManagement: auto
  # Default memory size for functions. Optional. (default: 1024MB).
  memorySize: 512
  # Default timeout for functions. Optional. (default: 6 seconds).
  # Note: API Gateway has a maximum timeout of 30 seconds
  timeout: 10
  # AWS Lambda Environment Variables for all functions. Optional.
  environment:
    APP_ENV_VARIABLE: FOOBAR
  # Duration for CloudWatch log retention. Optional. (default: forever).
  # Can be overridden for each function separately inside the functions block, see below on page.
  # Valid values: https://docs.aws.amazon.com/AWSCloudFormation/latest/UserGuide/aws-resource-logs-loggroup.html
  logRetentionInDays: 14
  # Policy defining how to monitor and mask sensitive data in CloudWatch logs. Optional.
  # Policy format: https://docs.aws.amazon.com/AmazonCloudWatch/latest/logs/mask-sensitive-log-data-start.html
  logDataProtectionPolicy:
    Name: data-protection-policy
  # KMS key ARN to use for encryption for all AWS Lambda functions. Optional.
  kmsKeyArn: arn:aws:kms:us-east-1:XXXXXX:key/some-hash
  # Version of hashing algorithm used by Serverless Framework for AWS Lambda function packaging. Optional.
  lambdaHashingVersion: 20201221
  # Use AWS Lambda function versioning. Optional. (enabled by default)
  versionFunctions: false
  # AWS Lambda Processor architecture: 'x86_64' or 'arm64' via Graviton2. Optional. (default: x86_64)
  architecture: x86_64

```

--------------------------------

### Customize CloudFormation resources in serverless.yml

Source: https://www.serverless.com/framework/docs/providers/aws/guide/serverless.yml

Customizes the CloudFormation template in serverless.yml to deploy extra resources. The example adds a DynamoDB table (usersTable), overrides the Properties of a Framework-created resource via the extensions section (UsersCreateLogGroup), and defines an Output (UsersTableArn) that exports the table ARN for cross-service references using Fn::ImportValue.

```yaml
# serverless.yml

# Insert raw CloudFormation (resources, outputs…) in the deployed template
resources:
  Resources:
    usersTable:
      Type: AWS::DynamoDB::Table
      Properties:
        TableName: usersTable
        AttributeDefinitions:
          - AttributeName: email
            AttributeType: S
        KeySchema:
          - AttributeName: email
            KeyType: HASH
        ProvisionedThroughput:
          ReadCapacityUnits: 1
          WriteCapacityUnits: 1

  extensions:
    # override Properties or other attributes of Framework-created resources.
    # See https://serverless.com/framework/docs/providers/aws/guide/resources#override-aws-cloudformation-resource for more details
    UsersCreateLogGroup:
      Properties:
        RetentionInDays: '30'

  # The "Outputs" that your AWS CloudFormation Stack should produce.  This allows references between services.
  Outputs:
    UsersTableArn:
      Description: The ARN for the User's Table
      Value: !GetAtt usersTable.Arn
      Export:
        # see Fn::ImportValue to use in other services
        # and http://docs.aws.amazon.com/AWSCloudFormation/latest/UserGuide/outputs-section-structure.html for documentation on use.
        Name: ${self:service}:${sls:stage}:UsersTableArn

```

--------------------------------

### Example config and resolved output

Source: https://www.serverless.com/framework/docs/providers/aws/cli-reference/print

Example serverless.yml with a variable reference, followed by the output of `sls print` showing the resolved values for `provider.stage` and `BucketName`.

```yaml
service: my-service

custom:
  bucketName: test

provider:
  name: aws
  runtime: nodejs14.x

functions:
  hello:
    handler: handler.hello

resources:
  Resources:
    MyBucket:
      Type: AWS::S3::Bucket
      Properties:
        BucketName: ${self:custom.bucketName}

```

```bash
$ sls print
service: my-service
custom:
  bucketName: test
provider:
  name: aws
  runtime: nodejs14.x
  stage: dev # <-- Resolved
functions:
  hello:
    handler: handler.hello
resources:
  Resources:
    MyBucket:
      Type: 'AWS::S3::Bucket'
      Properties:
        BucketName: test # <-- Resolved

```

--------------------------------

### package.json Dependencies for Code Interpreter

Source: https://www.serverless.com/framework/docs/providers/aws/guide/agents/code-interpreter

Required dependencies for the JavaScript code interpreter example. Install these packages to use the bedrock-agentcore SDK with LangChain and Zod.

```json
{
  "dependencies": {
    "bedrock-agentcore": "^0.2.0",
    "@langchain/aws": "^1.2.5",
    "@langchain/core": "^1.1.28",
    "langchain": "^1.2.27",
    "zod": "^3.23.0"
  }
}

```

--------------------------------

### Run serverless command to create a project

Source: https://www.serverless.com/framework/docs/examples-and-tutorials

Run this command to start the interactive Serverless Framework CLI and choose a built-in template from the list.

```bash
serverless

```

--------------------------------

### Install MCP SDK and Zod

Source: https://www.serverless.com/framework/docs/providers/aws/guide/mcp

Initializes a new Node.js project and installs the MCP server SDK and Zod for input validation. Run in a fresh directory before writing the server code.

```bash
npm init -y
npm install @modelcontextprotocol/server zod

```

--------------------------------

### Rule priority auto-assignment example

Source: https://www.serverless.com/framework/docs/providers/aws/guide/appsync/WAF

Shows how priorities are auto-assigned when not set, starting at 100 and incrementing sequentially for global rules then API key rules.

```yaml
appSync:
  waf:
    enabled: true
    rules:
      - name: Rule1
        # (no-set) Priority = 100
      - name: Rule2
        priority: 5 # Priority = 5
      - name: Rule3
        # (no-set) Priority = 101
  apiKeys:
    - name: Key1
      wafRules:
        - name: Rule4
          # (no-set) Priority = 102
        - name: Rule5
          # (no-set) Priority = 103
    - name: Key
      wafRules:
        - name: Rule6
          priority: 1 # Priority = 1
        - name: Rule7
          # (no-set) Priority = 104

```

--------------------------------

### External CloudFormation file structure

Source: https://www.serverless.com/framework/docs/guides/variables

Shows the required structure for external CloudFormation files: they must start with a Resources entity. This example defines an S3 bucket resource.

```yaml
Resources:
  Type: 'AWS::S3::Bucket'
  Properties:
    BucketName: some-bucket-name

```

--------------------------------

### samconfig.toml configuration example

Source: https://www.serverless.com/framework/docs/guides/sam

Example samconfig.toml file showing the parameters the Serverless Framework reads. Only the <stage>.global.parameters and <stage>.deploy.parameters sections are used, even for non-deploy commands.

```toml
version = 0.1

[default.deploy.parameters]
stack_name = "my-dev-stack"
region = "us-east-1"
template_file = "template.yml"
# s3_bucket = "my-bucket"
# parameter_overrides = "Environment=dev"

```

--------------------------------

### Add alarms block with notify set to SADN TOPIC ARN

Source: https://www.serverless.com/framework/docs/providers/aws/guide/sandboxes

Minimal configuration to enable CloudWatch alarms for a service's 'api' sandbox by specifying the SADN topic ARN under 'notify'.Defaults are used unless overridden: threshold 5, period  to 300 seconds evaluation period=1 datapoints=1 GreaterThanThreshold and missing data treated as notBreaching alarms are off by default and require this notify target; without notify the whole block is rejected by the framework. Note that the example intentionally omits preceding keys like 'observability' for brevity but implies they exist in the same file sandboxes declaration; nevertheless it is exactly as shown in the source: ```
``` and ```
``` and ```////////////// ``` and ``` and ``` and ``` and ``` and ``` and ``` and ``` and ``` and ``` and ``` and ``` and ``` and ``` and ``` and ``` and ``` and ``` and ``` and ``` and ``` and ``` and ``` and ``` and ``` and ``` and ``` and ``` and ``` snippet code is exactly as shown in source. Actual code snippet is exactly as shown in source: ```cpp
```cpp
```cpp
```cpp
```cpp
```cpp
```cpp
```cpp
```pid 
```cpp
```cpp
```cpp
```cpp
```cpp
```cpp
```cpp
```cpp
```cpp
```cpp
```cpp
```cpp
```cpp
```cpp
?{ 

```unknown - plaintext/config fence without language label; kept generic to avoid mislabeling code fence is exactly as in source: ```cpp
```cpp
```cpp
```cpp
```cpp
    } ] }, { 
sandboxes:
  api:
    artifact: ./app
    nesting: {}
    observability:
      alarms:
        notify: aris:aws:som:123456789012:my-alerts

```

--------------------------------

### Quick Start: Configure custom domain and deploy

Source: https://www.serverless.com/framework/docs/providers/aws/guide/domains

Configure a custom domain in the provider section and deploy. The framework automatically handles Route53 DNS, SSL certificate creation, and API Gateway domain mapping. After deployment, allow 2-5 minutes for DNS propagation and SSL activation.

```yaml
# serverless.yml
service: my-service

provider:
  name: aws
  runtime: nodejs20.x
  domain: api.example.com

functions:
  hello:
    handler: src/hello.handler
    events:
      - httpApi:
          path: /
          method: get

```

```bash
serverless deploy

```

--------------------------------

### Reference SNS topic defined in serverless.yml

Source: https://www.serverless.com/framework/docs/providers/aws/events/sns

References a topic defined in the same serverless.yml file using `!Ref` to get its ARN. Do not build a string as in the previous example when the topic is defined in the file.

```yaml
functions:
  dispatcher:
    handler: dispatcher.dispatch
    events:
      - sns:
          arn: !Ref SuperTopic
          topicName: MyCustomTopic

resources:
  Resources:
    SuperTopic:
      Type: AWS::SNS::Topic
      Properties:
        TopicName: MyCustomTopic

```

--------------------------------

### Session-based memory behavior example

Source: https://www.serverless.com/framework/docs/providers/aws/guide/agents/memory

Illustrates how memory retains conversation history per session ID. Requests with the same session ID can retrieve prior context, while a different session ID starts with no history.

```text
Request 1 (session: abc-123)
  User: "My name is Alice"
  Agent: "Nice to meet you, Alice!"
  → Stored in memory

Request 2 (session: abc-123)
  User: "What's my name?"
  Agent: "Your name is Alice"  ← Retrieved from memory

Request 3 (session: xyz-789)  # Different session
  User: "What's my name?"
  Agent: "I don't know your name yet"  ← No history for this session

```

--------------------------------

### Example project tree for unbundled build

Source: https://www.serverless.com/framework/docs/providers/aws/guide/building

Example project tree used to illustrate the unbundled build. Shows the source files and dependencies that will be packaged when bundle: false is set.

```text
serverless.yml
package.json          "dependencies": { "uuid": "^11.1.0" }, plus devDependencies
package-lock.json
tsconfig.json
.env
src/handler.ts        imports ./util and uuid, reads ../assets/banner.txt at runtime
src/util.ts
src/types.d.ts
assets/banner.txt
README.md

```

--------------------------------

### serverless onboarding app selection prompt

Source: https://www.serverless.com/framework/docs/getting-started

Shown during the `serverless` onboarding command when setting up an App. Select an existing app, create a new one, or skip adding an app.

```text
❯ Create A New App
  ecommerce
  blog
  acmeinc
  Skip Adding An App

```
Tokens:
Raw