Upload to dashboard

After a CLI scan, you can send dataflow.json to DataParade and open Preview & Edit in the web app — without manually uploading the file through the create dialog.

The upload creates an import preview draft (a saved preview session), not a finished assessment. You review components, flows, and properties in the app, then create the diagram when ready.

Auto-upload after scan

After every successful scan (unless you opt out), the CLI auto-uploads dataflow.json to DataParade and prints a Preview ready URL.

Opt out: --skip-auto-upload or DATAPARADE_SKIP_AUTO_UPLOAD=true / 1
With workspace API key (DATAPARADE_WORKSPACE_API_KEY from Workspace → Access keys): preview opens on the dashboard for that workspace (?importDraft=)
Without a workspace API key: preview is stored temporarily; the URL opens sign up (/preview/cli/<token>). After you create an account, the scan appears on your dashboard.
Upload alone does not consume scan quota — only hosted scans and CLI runs with platform AI use workspace scan quotas

Console output with workspace API key:


[scan] dataflow.json written to /path/to/dataflow.json
[scan] Preview ready: https://app.dataparade.io/dashboard?importDraft=<draftId>

Console output without workspace API key:


[scan] dataflow.json written to /path/to/dataflow.json
[scan] Preview ready: https://app.dataparade.io/preview/cli/<claimToken>
[scan] Claiming this preview after signup uses 1 scan slot.

Auto-upload uses the same assessment name as the scan’s --project-name / projectName in config when set. Otherwise the preview defaults to Imported scan YYYY-MM-DD.

`upload` command

Upload an existing dataflow.json without re-scanning:


npx @dataparade/cli upload ./dataflow.json --project-name "My service"

Argument / option	Purpose
`<file>`	Path to a valid `dataflow.json` wrapper (see Output and results)
`--project-name <name>`	Assessment name shown in preview (optional; falls back to `metadata.projectName` in the file, then `Imported scan YYYY-MM-DD`)
`--workspace-api-key <key>`	Workspace API key (or set `DATAPARADE_WORKSPACE_API_KEY`). Optional — without a key, upload uses the anonymous preview + signup flow.

On success the CLI prints:


[upload] Preview ready: https://app.dataparade.io/dashboard?importDraft=<draftId>

Preview deep link

Workspace API key: the URL opens the dashboard with ?importDraft=<draftId>. If you are signed in to the matching workspace, the CLI import preview modal opens automatically.

No API key: the URL opens /preview/cli/<claimToken>, redirects you to sign up, then attaches the preview to your new account and opens the same import modal on the dashboard.

Confirm or edit the assessment name (required before creating the diagram).
Review assets, third parties, actors, and data flows — same preview as manual CLI import.
Create the assessment when satisfied.

Workspace owners and editors can open drafts uploaded to that workspace. Viewers cannot create assessments from CLI uploads.

Access keys

Workspace owners create CLI access keys on Workspace → Access keys. The same key authenticates:

Upload and auto-upload (POST /api/scans/cli/upload with a key, or POST /api/scans/cli/preview without)
Platform AI when inference is on (default): quota preflight, infer tasks, complete
- Anonymous scans: session + IP caps; actual tokens accrue on the job and debit at claim. The ~100k figure is a server hard ceiling, not the working scan budget (CLI defaults to ~12k unless you raise SCAN_AI_BUDGET_TOKENS)

See Workspace and team for key limits and handling.

Environment variables

Variable	Purpose
`DATAPARADE_WORKSPACE_API_KEY`	Auth for upload and platform AI
`DATAPARADE_API_BASE_URL`	Backend API origin (omit for published npm CLI; use `http://localhost:3000` for local backend)
`DATAPARADE_APP_URL`	Frontend origin for preview links (omit for production; use `http://localhost:3001` for local frontend)
`DATAPARADE_SKIP_AUTO_UPLOAD`	Skip post-scan upload when `true` or `1`

Full detail: Environment variables.

Quota

Action	Uses scan quota?
Structural-only `scan` (no AI)	No — even with a workspace key
`upload` / auto-upload	No
`scan` with platform AI (default; workspace key)	Yes — preflight, per-infer billing, complete
`scan` with anonymous platform AI (no workspace key)	Session + IP cap; tokens debit on claim

See Workspace scan quotas.

CLI Run and Export — build and run scan
Creation and Imports — manual file import and preview behavior
Output and results — dataflow.json format and console lines
Troubleshooting — upload and import errors