- Home
- Skills
- Data & Databases
- Salesforce Data 360 SFTP Ingestion Pro
Works with the AI Agents you already use
Salesforce Data 360 SFTP Ingestion Pro
Validate Parquet schemas and publish files to Salesforce Data 360 via guarded SFTP staging scripts.
$9
Salesforce Data 360 SFTP Ingestion Pro
Example session with this skill installed
Validate my orders.parquet file against the orders-contract.json schema and publish it to the /salesforce/inbound directory.
- Read your context and instructions
- Compiled the salesforce data 360
I have completed the validation and publication process for the orders data. The file passed all schema checks and has been staged on the remote SFTP server.
Publication Audit Report
| Task | Status | Details |
|---|---|---|
| Schema Validation | PASS | All 14 fields match orders-contract.json |
| Key Check | PASS | 0 nulls, 0 duplicates found in order_id |
| Row Count | 12,500 | Matches expected batch size |
| Staging | SUCCESS | Uploaded as orders.parquet.part |
| Final Rename | SUCCESS | Atomic rename to orders_20231027_T1430.parquet |
| Remote Path | COMPLETED | /salesforce/inbound/orders_20231027_T1430.parquet |
{
"event": "publication_success",
"timestamp": "2023-10-27T14:30:05Z",
"contract": "orders-contract.json",
"remote_file": "orders_20231027_T1430.parquet",
"stats": {
"rows": 12500,
"errors": 0
}
}
Next steps
- Log in to your Salesforce Data 360 org and navigate to Data Streams.
- Refresh the stream associated with the
/salesforce/inbounddirectory to begin ingestion. - Check the Data Lake Object (DLO) metadata to verify the new records are visible.
Connects securely to your tools. The creator never sees your data.
About this skill
Manual SFTP uploads to Salesforce Data 360 often lead to ingestion failures due to schema drift or primary key violations. This skill enforces data integrity by validating Parquet files against local contracts before they hit the staging server.
What it does
- Contract validation verifies Parquet schemas against JSON definitions to prevent upstream ingestion errors.
- Key integrity checks identifies duplicate or null primary keys and validates record grain before publication.
- Guarded SFTP staging uploads files using
.partextensions and atomic renames to ensure Data 360 doesn't pick up partial files. - Unique file naming generates compliant, unique filenames to prevent accidental overwrites in the staging directory.
- Lifecycle auditing generates JSON reports for every validation and publication step to provide an audit trail of source quality.
How it works
- Define the contract by specifying the physical schema, primary keys, and refresh semantics in a JSON manifest.
- Execute validation using the provided scripts to check Parquet files for schema compliance and key quality.
- Stage for publication by configuring environment variables for host, user, and SSH keys.
- Publish to SFTP using the guarded script which handles the staging, renaming, and remote path recording.
- Monitor Ingestion by verifying the resulting Data Stream runs and DLO freshness within Salesforce.
Frameworks & tools
This tool utilizes Node.js for validation logic and OpenSSH sftp for secure file transfers. It is designed for workflows targeting Salesforce Data 360 (Data Cloud).
Why this beats prompting it yourself
Prompting an agent to "upload a file" ignores the specific atomic renaming requirements and schema strictness of Data 360. This skill provides the exact shell scripts and validation logic needed to handle Parquet-to-D360 pipelines without manual intervention or corrupt data streams.
Use cases
- Pre-flight validation for automated ETL pipelines pushing Parquet files to Salesforce.
- Debugging schema drift when Data 360 streams fail due to unexpected field changes.
- Key quality audits to ensure upsert keys are unique and non-null before ingestion.
Known limitations
Requires local OpenSSH and Node.js environments. Does not include Data 360 API credentials or scheduling logic.
Requires local Node.js and OpenSSH sftp client. Does not trigger Data 360 ingestion internally; it only manages the publication to the staging area.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 5 days ago
Requires
Permissions
Allowed Hosts