Credentials

Securely manage external service credentials and integrations for use in your API flows and edge functions.

Overview

The Credentials section lets you connect your project to external services like AWS, Google, databases, and messaging platforms. All credentials are encrypted with AES-256 at rest and are never exposed in the UI after being saved.

Once stored, credentials can be referenced from two places:

• Flow Editor — Use the Credentials Select widget when configuring nodes
• Edge Functions — Access credentials via the dypai.credentials module

Supported Integrations

Adding Credentials

Using Credentials in Flows

When configuring a node in the Flow Editor that requires external service access, use the Credentials Select widget to pick from your saved credentials. The credential is automatically injected at execution time.

Using Credentials in Edge Functions

Access stored credentials programmatically from Python or Rust edge functions:

# Python Edge Function
async def handler(request):
    # Retrieve a stored credential by name
    aws = dypai.credentials.get("my-aws-credential")

    # Use the credential fields
    s3_client = boto3.client(
        's3',
        aws_access_key_id=aws["access_key"],
        aws_secret_access_key=aws["secret_key"],
        region_name=aws["region"]
    )

    buckets = s3_client.list_buckets()
    return {"buckets": buckets}

Security

Managing Credentials

From the credentials list you can:

• Update — Re-enter credential values to rotate keys or update tokens
• Delete — Remove a credential permanently

Next steps

• Use credentials in the Flow Editor
• Access credentials from Edge Functions
• Manage environment variables