Cloud linking
Connect a cloud account to your Ductape workspace once. Reference it by tag on any product component. Ductape imports or provisions the resource, stores credentials as workspace secrets, and refreshes them at runtime.
Supported providers: AWS, GCP, Azure, MongoDB Atlas, Neo4j Aura.
How it works
| Step | What you do | Docs |
|---|---|---|
| 1. Connect | Create a workspace cloud connection, add provider credentials, validate | Cloud connections |
| 2. Link | Set cloud: '<tag>' and a resource name on a component env | Link components |
| 3. Run | Use ductape.databases.connect(), ductape.storage.upload(), etc. — credentials resolve automatically | Component docs below |
Connection tags
Every connection has a stable tag (for example prod_aws, prod_gcp). Use that tag everywhere — never an internal connection ID.
- TypeScript
- Java
- Go
- .NET
// On a database env
{ slug: 'prd', cloud: 'prod_aws', instance: 'my-app-db', region: 'us-east-1' }
// On low-level cloud APIs
await ductape.cloud.resources.list({ cloud: 'prod_aws', service: 's3', region: 'us-east-1' });
// On a database env
Map.of( "slug", "prd", "cloud", "prod_aws", "instance", "my-app-db", "region", "us-east-1" )
// On low-level cloud APIs
ductape.cloud.resources.list(Map.of( "cloud", "prod_aws", "service", "s3", "region", "us-east-1" ));
// On a database env
{ "slug": "prd", "cloud": "prod_aws", "instance": "my-app-db", "region": "us-east-1" }
// On low-level cloud APIs
client.cloud.resources.list({ "cloud": "prod_aws", "service": "s3", "region": "us-east-1" });
// On a database env
{ ["slug"] = "prd", ["cloud"] = "prod_aws", ["instance"] = "my-app-db", ["region"] = "us-east-1" }
// On low-level cloud APIs
await ductape.cloud.resources.list({ ["cloud"] = "prod_aws", ["service"] = "s3", ["region"] = "us-east-1" });
What you can link
| Component | AWS | GCP | Azure | Atlas | Aura |
|---|---|---|---|---|---|
| Storage | S3 | GCS | Blob | — | — |
| Message brokers | SQS | Pub/Sub | Service Bus | — | — |
| Databases | RDS (PostgreSQL) | Cloud SQL | PostgreSQL Flexible | MongoDB cluster | — |
| Graphs | Neptune | Spanner Graph | Cosmos Gremlin | — | Aura instance |
| Vectors | OpenSearch | Vertex Vector Search | AI Search | — | — |
Provider setup and IAM/RBAC checklists live on each provider page.
Workbench or SDK
Both paths do the same thing:
- Workbench — Cloud sidebar → add connection → link from the component editor when you save.
- SDK —
connections.create→connections.complete→ create/update component withcloudon the env.
Next
- Set up a cloud connection
- Link your first component
- AWS networking — only if you use RDS or Neptune