View a markdown version of this page

Using the AWS Secrets Manager Agent - AWS Secrets Manager
How the Secrets Manager Agent worksUnderstanding Secrets Manager Agent cachingBuild the Secrets Manager AgentInstall the Secrets Manager AgentRetrieve secrets with the Secrets Manager AgentUnderstanding the refreshNow parameterRetrieve secrets across accounts with role chainingPre-fetch secrets at startupConfiguration optionsOptional featuresLoggingSecurity considerations

Using the AWS Secrets Manager Agent

How the Secrets Manager Agent works

The AWS Secrets Manager Agent is a client-side HTTP service that helps you standardize how you consume secrets from Secrets Manager across your compute environments. You can use it with the following services:

  • AWS Lambda

  • Amazon Elastic Container Service

  • Amazon Elastic Kubernetes Service

  • Amazon Elastic Compute Cloud

The Secrets Manager Agent retrieves and caches secrets in memory, allowing your applications to get secrets from localhost instead of making direct calls to Secrets Manager. The Secrets Manager Agent can only read secrets—it can't modify them.

The Secrets Manager Agent is open source. The source code, installation instructions, and latest release information are available on GitHub.

Important

The Secrets Manager Agent uses the AWS credentials from your environment to call Secrets Manager. It includes protection against Server Side Request Forgery (SSRF) to help improve secret security. The Secrets Manager Agent uses the post-quantum ML-KEM key exchange as the highest-priority key exchange by default.

Understanding Secrets Manager Agent caching

The Secrets Manager Agent uses an in-memory cache that resets when the Secrets Manager Agent restarts. It periodically refreshes cached secret values based on the following:

  • The default refresh frequency (TTL) is 300 seconds

  • You can modify the TTL using a configuration file

  • The refresh occurs when you request a secret after the TTL expires

Note

The Secrets Manager Agent doesn't include cache invalidation. If a secret rotates before the cache entry expires, the Secrets Manager Agent might return a stale secret value.

The Secrets Manager Agent returns secret values in the same format as the response of GetSecretValue. Secret values aren't encrypted in the cache.

Build the Secrets Manager Agent

Before you begin, ensure you have the standard development tools and Rust tools installed for your platform.

Note

Building the agent with the fips feature enabled on macOS currently requires the following workaround:

  • Create an environment variable called SDKROOT which is set to the result of running xcrun --show-sdk-path

RPM-based systems
To build on RPM-based systems

  1. Use the install script provided in the repository.

    The script generates a random SSRF token on startup and stores it in the file /var/run/awssmatoken. The token is readable by the awssmatokenreader group that the install script creates.

  2. To allow your application to read the token file, you need to add the user account that your application runs under to the awssmatokenreader group. For example, you can grant permissions for your application to read the token file with the following usermod command, where <APP_USER> is the user ID under which your application runs.

    sudo usermod -aG awssmatokenreader <APP_USER>
    Install development tools

    On RPM-based systems such as AL2023, install the Development Tools group:

    sudo yum -y groupinstall "Development Tools"
  3. Install Rust

    Follow the instructions at Install Rust in the Rust documentation:

    curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh # Follow the on-screen instructions
. "$HOME/.cargo/env"
  4. Build the agent

    Build the Secrets Manager Agent using the cargo build command:

    cargo build --release

    You will find the executable under target/release/aws_secretsmanager_agent.

Debian-based systems
To build on Debian-based systems
  1. Install development tools

    On Debian-based systems such as Ubuntu, install the build-essential package:

    sudo apt install build-essential
  2. Install Rust

    Follow the instructions at Install Rust in the Rust documentation:

    curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh # Follow the on-screen instructions
. "$HOME/.cargo/env"
  3. Build the agent

    Build the Secrets Manager Agent using the cargo build command:

    cargo build --release

    You will find the executable under target/release/aws_secretsmanager_agent.

Windows
To build on Windows
  1. Set up development environment

    Follow the instructions at Set up your dev environment on Windows for Rust in the Microsoft Windows documentation.

  2. Build the agent

    Build the Secrets Manager Agent using the cargo build command:

    cargo build --release

    You will find the executable under target/release/aws_secretsmanager_agent.exe.

Cross-compile natively
To cross-compile natively
  1. Install cross-compile tools

    On distributions where the mingw-w64 package is available such as Ubuntu, install the cross-compile toolchain:

    # Install the cross compile tool chain
sudo add-apt-repository universe
sudo apt install -y mingw-w64
  2. Add Rust build targets

    Install the Windows GNU build target:

    rustup target add x86_64-pc-windows-gnu
  3. Build for Windows

    Cross-compile the agent for Windows:

    cargo build --release --target x86_64-pc-windows-gnu

    You will find the executable at target/x86_64-pc-windows-gnu/release/aws_secretsmanager_agent.exe.

Cross compile with Rust cross
To cross-compile using Rust cross

If the cross-compile tools are not available natively on the system, you can use the Rust cross project. For more information, see https://github.com/cross-rs/cross.

Important

We recommend 32GB disk space for the build environment.

  1. Set up Docker

    Install and configure Docker:

    # Install and start docker
sudo yum -y install docker
sudo systemctl start docker
sudo systemctl enable docker # Make docker start after reboot
  2. Configure Docker permissions

    Add your user to the docker group:

    # Give ourselves permission to run the docker images without sudo
sudo usermod -aG docker $USER
newgrp docker
  3. Build for Windows

    Install cross and build the executable:

    # Install cross and cross compile the executable
cargo install cross
cross build --release --target x86_64-pc-windows-gnu

Install the Secrets Manager Agent

Choose your compute environment from the following installation options.

Amazon EC2
To install the Secrets Manager Agent on Amazon EC2
  1. Navigate to configuration directory

    Change to the configuration directory:

    cd aws_secretsmanager_agent/configuration
  2. Run installation script

    Run the install script provided in the repository.

    The script generates a random SSRF token on startup and stores it in the file /var/run/awssmatoken. The token is readable by the awssmatokenreader group that the install script creates.

  3. Configure application permissions

    Add the user account that your application runs under to the awssmatokenreader group:

    sudo usermod -aG awssmatokenreader APP_USER

    Replace APP_USER with the user ID under which your application runs.

Container Sidecar

You can run the Secrets Manager Agent as a sidecar container alongside your application by using Docker. Then your application can retrieve secrets from the local HTTP server the Secrets Manager Agent provides. For information about Docker, see the Docker documentation.

To create a sidecar container for the Secrets Manager Agent
  1. Create agent Dockerfile

    Create a Dockerfile for the Secrets Manager Agent sidecar container:

    # Use the latest Debian image as the base
FROM debian:latest

# Set the working directory inside the container
WORKDIR /app 

# Copy the Secrets Manager Agent binary to the container
COPY secrets-manager-agent . 

# Install any necessary dependencies
RUN apt-get update && apt-get install -y ca-certificates 

# Set the entry point to run the Secrets Manager Agent binary
ENTRYPOINT ["./secrets-manager-agent"]
  2. Create application Dockerfile

    Create a Dockerfile for your client application.

  3. Create Docker Compose file

    Create a Docker Compose file to run both containers with a shared network interface:

    Important

    You must load AWS credentials and the SSRF token for the application to be able to use the Secrets Manager Agent. For Amazon EKS and Amazon ECS, see the following:

    version: '3'
services:
    client-application:
    container_name: client-application
    build:
        context: .
        dockerfile: Dockerfile.client
    command: tail -f /dev/null  # Keep the container running
    

    secrets-manager-agent:
    container_name: secrets-manager-agent
    build:
        context: .
        dockerfile: Dockerfile.agent
    network_mode: "container:client-application"  # Attach to the client-application container's network
    depends_on:
        - client-application
  4. Copy agent binary

    Copy the secrets-manager-agent binary to the same directory that contains your Dockerfiles and Docker Compose file.

  5. Build and run containers

    Build and run the containers using Docker Compose:

    docker-compose up --build
  6. Next steps

    You can now use the Secrets Manager Agent to retrieve secrets from your client container. For more information, see Retrieve secrets with the Secrets Manager Agent.

Lambda

You can package the Secrets Manager Agent as a Lambda extension. Then you can add it to your Lambda function as a layer and call the Secrets Manager Agent from your Lambda function to get secrets.

The following instructions show how to get a secret named MyTest by using the example script secrets-manager-agent-extension.sh in https://github.com/aws/aws-secretsmanager-agent to install the Secrets Manager Agent as a Lambda extension.

To create a Lambda extension for the Secrets Manager Agent
  1. Package the agent layer

    From the root of the Secrets Manager Agent code package, run the following commands:

    AWS_ACCOUNT_ID=AWS_ACCOUNT_ID
LAMBDA_ARN=LAMBDA_ARN

# Build the release binary 
cargo build --release --target=x86_64-unknown-linux-gnu

# Copy the release binary into the `bin` folder
mkdir -p ./bin
cp ./target/x86_64-unknown-linux-gnu/release/aws_secretsmanager_agent ./bin/secrets-manager-agent

# Copy the `secrets-manager-agent-extension.sh` example script into the `extensions` folder.
mkdir -p ./extensions
cp aws_secretsmanager_agent/examples/example-lambda-extension/secrets-manager-agent-extension.sh ./extensions

# Zip the extension shell script and the binary 
zip secrets-manager-agent-extension.zip bin/* extensions/*

# Publish the layer version
LAYER_VERSION_ARN=$(aws lambda publish-layer-version \
    --layer-name secrets-manager-agent-extension \
    --zip-file "fileb://secrets-manager-agent-extension.zip" | jq -r '.LayerVersionArn')
  2. Configure SSRF token

    The default configuration of the agent will automatically set the SSRF token to the value set in the pre-set AWS_SESSION_TOKEN or AWS_CONTAINER_AUTHORIZATION_TOKEN environment variables (the latter variable for Lambda functions with SnapStart enabled). Alternatively, you can define the AWS_TOKEN environment variable with an arbitrary value for your Lambda function instead as this variable takes precedence over the other two. If you choose to use the AWS_TOKEN environment variable, you must set that environment variable with a lambda:UpdateFunctionConfiguration call.

  3. Attach layer to function

    Attach the layer version to your Lambda function:

    # Attach the layer version to the Lambda function
aws lambda update-function-configuration \
    --function-name $LAMBDA_ARN \
    --layers "$LAYER_VERSION_ARN"
  4. Update function code

    Update your Lambda function to query http://localhost:2773/secretsmanager/get?secretId=MyTest with the X-Aws-codes-Secrets-Token header value set to the value of the SSRF token sourced from one the environment variables mentioned above to retrieve the secret. Be sure to implement retry logic in your application code to accommodate delays in initialization and registration of the Lambda extension.

  5. Test the function

    Invoke the Lambda function to verify that the secret is being correctly fetched.

Retrieve secrets with the Secrets Manager Agent

To retrieve a secret, call the local Secrets Manager Agent endpoint with the secret name or ARN as a query parameter. By default, the Secrets Manager Agent retrieves the AWSCURRENT version of the secret. To retrieve a different version, use either the versionStage or versionId parameter.

Important

To help protect the Secrets Manager Agent, you must include a SSRF token header as part of each request: X-Aws-Parameters-Secrets-Token. The Secrets Manager Agent denies requests that don't have this header or that have an invalid SSRF token. You can customize the SSRF header name in the Configure the Secrets Manager Agent.

Required permissions

The Secrets Manager Agent uses the AWS SDK for Rust, which uses the AWS credential provider chain. The identity of these IAM credentials determines the permissions the Secrets Manager Agent has to retrieve secrets.

  • secretsmanager:DescribeSecret

  • secretsmanager:GetSecretValue

For more information about permissions, see Permissions reference for AWS Secrets Manager.

Important

After the secret value is pulled into the Secrets Manager Agent, any user with access to the compute environment and SSRF token can access the secret from the Secrets Manager Agent cache. For more information, see Security considerations.

Example requests

curl
Example – Get a secret using curl

The following curl example shows how to get a secret from the Secrets Manager Agent. The example relies on the SSRF being present in a file, which is where it is stored by the install script.

curl -v -H \
    "X-Aws-Parameters-Secrets-Token: $(</var/run/awssmatoken)" \
    'http://localhost:2773/secretsmanager/get?secretId=YOUR_SECRET_ID'
Python
Example – Get a secret using Python

The following Python example shows how to get a secret from the Secrets Manager Agent. The example relies on the SSRF being present in a file, which is where it is stored by the install script.

import requests
import json

# Function that fetches the secret from Secrets Manager Agent for the provided secret id. 
def get_secret():
    # Construct the URL for the GET request
    url = f"http://localhost:2773/secretsmanager/get?secretId=YOUR_SECRET_ID"

    # Get the SSRF token from the token file
    with open('/var/run/awssmatoken') as fp:
        token = fp.read() 

    headers = {
        "X-Aws-Parameters-Secrets-Token": token.strip()
    }

    try:
        # Send the GET request with headers
        response = requests.get(url, headers=headers)

        # Check if the request was successful
        if response.status_code == 200:
            # Return the secret value
            return response.text
        else:
            # Handle error cases
            raise Exception(f"Status code {response.status_code} - {response.text}")

    except Exception as e:
        # Handle network errors
        raise Exception(f"Error: {e}")

Understanding the refreshNow parameter

The Secrets Manager Agent uses an in-memory cache to store secret values, which it refreshes periodically. By default, this refresh occurs when you request a secret after the Time to Live (TTL) has expired, typically every 300 seconds. However, this approach can sometimes result in stale secret values, especially if a secret rotates before the cache entry expires.

To address this limitation, the Secrets Manager Agent supports a parameter called refreshNow in the URL. You can use this parameter to force an immediate refresh of a secret's value, bypassing the cache and ensuring you have the most up-to-date information.

Default behavior (without refreshNow)

  • Uses cached values until TTL expires

  • Refreshes secrets only after TTL (default 300 seconds)

  • May return stale values if secrets rotate before the cache expires

Behavior with refreshNow=true

  • Bypasses the cache entirely

  • Retrieves the latest secret value directly from Secrets Manager

  • Updates the cache with the fresh value and resets the TTL

  • Ensures you always get the most current secret value

Force-refresh a secret value

Important

The default value of refreshNow is false. When set to true, it overrides the TTL specified in the Secrets Manager Agent configuration file and makes an API call to Secrets Manager.

curl
Example – Force-refresh a secret using curl

The following curl example shows how to force the Secrets Manager Agent to refresh the secret. The example relies on the SSRF being present in a file, which is where it is stored by the install script.

curl -v -H \
"X-Aws-Parameters-Secrets-Token: $(</var/run/awssmatoken)" \
'http://localhost:2773/secretsmanager/get?secretId=YOUR_SECRET_ID&refreshNow=true'
Python
Example – Force-refresh a secret using Python

The following Python example shows how to get a secret from the Secrets Manager Agent. The example relies on the SSRF being present in a file, which is where it is stored by the install script.

import requests
import json

# Function that fetches the secret from Secrets Manager Agent for the provided secret id. 
def get_secret():
    # Construct the URL for the GET request
    url = f"http://localhost:2773/secretsmanager/get?secretId=YOUR_SECRET_ID&refreshNow=true"

    # Get the SSRF token from the token file
    with open('/var/run/awssmatoken') as fp:
        token = fp.read() 

    headers = {
        "X-Aws-Parameters-Secrets-Token": token.strip()
    }

    try:
        # Send the GET request with headers
        response = requests.get(url, headers=headers)

        # Check if the request was successful
        if response.status_code == 200:
            # Return the secret value
            return response.text
        else:
            # Handle error cases
            raise Exception(f"Status code {response.status_code} - {response.text}")

    except Exception as e:
        # Handle network errors
        raise Exception(f"Error: {e}")

Retrieve secrets across accounts with role chaining

Role chaining enables the Secrets Manager Agent to retrieve secrets from other AWS accounts by assuming IAM roles using AWS STS AssumeRole. The Secrets Manager Agent creates and caches a separate caching client for each unique role ARN. Each role client maintains its own independent cache, so the same secret fetched with different roles has separate cache entries.

Required permissions

To use role chaining, you need the following:

  • The Secrets Manager Agent's environment credentials must have sts:AssumeRole permission on the target role ARN.

  • The target role must have secretsmanager:GetSecretValue and secretsmanager:DescribeSecret permissions for the secrets you want to access.

  • The target role's trust policy must allow the Secrets Manager Agent's identity to assume it.

Retrieve cross-account secrets

Include the roleArn query parameter in your request to the Secrets Manager Agent to specify which role to assume for the secret retrieval.

curl
Example – Cross-account secret using curl
curl -v -H \
    "X-Aws-Parameters-Secrets-Token: $(</var/run/awssmatoken)" \
    'http://localhost:2773/secretsmanager/get?secretId=YOUR_SECRET_ID&roleArn=arn:aws:iam::ACCOUNT_ID:role/ROLE_NAME'
Python
Example – Cross-account secret using Python
import requests

def get_secret_cross_account():
    secret_id = "YOUR_SECRET_ID"
    role_arn = "arn:aws:iam::ACCOUNT_ID:role/ROLE_NAME"
    url = f"http://localhost:2773/secretsmanager/get?secretId={secret_id}&roleArn={role_arn}"

    with open('/var/run/awssmatoken') as fp:
        token = fp.read()

    headers = {
        "X-Aws-Parameters-Secrets-Token": token.strip()
    }

    try:
        response = requests.get(url, headers=headers)

        if response.status_code == 200:
            return response.text
        else:
            raise Exception(f"Status code {response.status_code} - {response.text}")

    except Exception as e:
        raise Exception(f"Error: {e}")

Role chaining configuration and limits

Configure role chaining with the max_roles option in your TOML configuration file. This sets the maximum number of simultaneous assumed roles, in the range 1 to 20. The default is 20.

Important

Assumed roles are not evicted from the Secrets Manager Agent's role cache. Once the maximum number of roles has been reached, requests with new role ARNs are rejected with a 400 error until the Secrets Manager Agent is restarted.

Error responses for role chaining
400

The roleArn format is invalid or the maximum number of assumed roles has been reached.

403

The AWS STS AssumeRole call failed. Verify that the target role's trust policy allows the Secrets Manager Agent's identity to assume it.

Pre-fetch secrets at startup

By default, the Secrets Manager Agent fetches secrets on demand when your application requests them. With pre-fetching, the Secrets Manager Agent loads specified secrets into the cache when it starts up, so your application can access them immediately without waiting for the first API call. Pre-fetching runs as a background task—the Secrets Manager Agent begins accepting requests immediately and does not block on pre-fetch completion.

You can specify secrets to pre-fetch in two ways:

  • Explicit secrets – List specific secret IDs or ARNs.

  • Tag-based discovery – Discover secrets by tag key. The Secrets Manager Agent fetches all secrets that have the specified tag.

Required permissions

In addition to the standard permissions for retrieving secrets, pre-fetching requires the following:

  • secretsmanager:BatchGetSecretValue – Required for all pre-fetch operations.

  • secretsmanager:ListSecrets – Required only when using tag-based discovery.

Configure pre-fetching

Add a [prefetch] section to your TOML configuration file. The following options are available:

cache_buffer_ratio

The maximum fraction of the cache to fill per client during pre-fetch, in the range 0.1 to 1.0. The default is 0.8. When the buffer limit is reached, the Secrets Manager Agent stops pre-fetching remaining secrets—it does not evict existing cache entries. Secrets not loaded during pre-fetch are still available on demand.

max_jitter_seconds

A random delay in seconds before pre-fetching begins, in the range 0 to 10. The default is 0. Use this to prevent synchronized fleet-wide API calls when multiple agents start at the same time.

Example Pre-fetch configuration with explicit secrets
[prefetch]
cache_buffer_ratio = 0.6
max_jitter_seconds = 5
secrets = [
    { secret_id = "arn:aws:secretsmanager:us-west-2:123456789012:secret:MySecret-AbCdEf" },
    { secret_id = "MyOtherSecret" },
]
Example Pre-fetch configuration with tag-based discovery
[prefetch]
cache_buffer_ratio = 0.8
filter_tags = [
    { key = "Environment" },
    { key = "Team" },
]

You can also combine explicit secrets and tag-based discovery in the same configuration. For cross-account pre-fetching, add the role_arn field. For more information, see Retrieve secrets across accounts with role chaining.

Example Pre-fetch configuration with cross-account access
[prefetch]
cache_buffer_ratio = 0.6
max_jitter_seconds = 5
secrets = [
    { secret_id = "arn:aws:secretsmanager:us-west-2:123456789012:secret:MySecret-AbCdEf" },
    { secret_id = "cross-account-secret", role_arn = "arn:aws:iam::987654321098:role/SecretAccessRole" },
]
filter_tags = [
    { key = "Environment" },
    { key = "Team", role_arn = "arn:aws:iam::987654321098:role/SecretAccessRole" },
]

Configure the Secrets Manager Agent

To change the configuration of the Secrets Manager Agent, create a TOML config file, and then call ./aws_secretsmanager_agent --config config.toml.

Configuration options
log_level

The level of detail reported in logs for the Secrets Manager Agent: DEBUG, INFO, WARN, ERROR, or NONE. The default is INFO.

log_to_file

Whether to log to a file or stdout/stderr: true or false. The default is true.

http_port

The port for the local HTTP server, in the range 1024 to 65535. The default is 2773.

region

The AWS Region to use for requests. If no Region is specified, the Secrets Manager Agent determines the Region from the SDK. For more information, see Specify your credentials and default Region in the AWS SDK for Rust Developer Guide.

ttl_seconds

The TTL in seconds for the cached items, in the range 0 to 3600. The default is 300. 0 indicates that there is no caching.

cache_size

The maximum number of secrets that can be stored in the cache, in the range 1 to 1000. The default is 1000.

ssrf_headers

A list of header names the Secrets Manager Agent checks for the SSRF token. The default is "X-Aws-Parameters-Secrets-Token, X-Vault-Token".

ssrf_env_variables

A list of environment variable names the Secrets Manager Agent checks in sequential order for the SSRF token. The environment variable can contain the token or a reference to the token file as in: AWS_TOKEN=file:///var/run/awssmatoken. The default is "AWS_TOKEN, AWS_SESSION_TOKEN, AWS_CONTAINER_AUTHORIZATION_TOKEN".

path_prefix

The URI prefix used to determine if the request is a path based request. The default is "/v1/".

max_conn

The maximum number of connections from HTTP clients that the Secrets Manager Agent allows, in the range 1 to 1000. The default is 800.

max_roles

The maximum number of simultaneous IAM roles for cross-account access, in the range 1 to 20. The default is 20. For more information, see Retrieve secrets across accounts with role chaining.

Optional features

The Secrets Manager Agent can be built with optional features by passing the --features flag to cargo build. The available features are:

Build features
prefer-post-quantum

Makes X25519MLKEM768 the highest-priority key exchange algorithm. Otherwise, it is available but not highest-priority. X25519MLKEM768 is a hybrid, post-quantum-secure key exchange algorithm.

fips

Restricts the cipher suites used by the agent to only FIPS-approved ciphers.

Logging

Local logging

The Secrets Manager Agent logs errors locally to the file logs/secrets_manager_agent.log or to stdout/stderr depending on the log_to_file config variable. When your application calls the Secrets Manager Agent to get a secret, those calls appear in the local log. They do not appear in the CloudTrail logs.

Log rotation

The Secrets Manager Agent creates a new log file when the file reaches 10 MB, and it stores up to five log files total.

AWS service logging

The log does not go to Secrets Manager, CloudTrail, or CloudWatch. Requests to get secrets from the Secrets Manager Agent do not appear in those logs. When the Secrets Manager Agent makes a call to Secrets Manager to get a secret, that call is recorded in CloudTrail with a user agent string containing aws-secrets-manager-agent.

You can configure logging options in the Configure the Secrets Manager Agent.

Security considerations

Domain of trust

For an agent architecture, the domain of trust is where the agent endpoint and SSRF token are accessible, which is usually the entire host. The domain of trust for the Secrets Manager Agent should match the domain where the Secrets Manager credentials are available in order to maintain the same security posture. For example, on Amazon EC2 the domain of trust for the Secrets Manager Agent would be the same as the domain of the credentials when using roles for Amazon EC2.

Important

Security conscious applications that are not already using an agent solution with the Secrets Manager credentials locked down to the application should consider using the language-specific AWS SDKs or caching solutions. For more information, see Get secrets.