dlt-hub Deploy pipeline.py — already present in the project root — to dltHub Platform. This pipeline loads data from the Jaffle Shop API into the dltHub playground cloud data warehouse (cloud storage handled by dltHub — no credentials needed).
Do not use when pipeline.py does not exist in the project root.
If the user wants to build their own pipeline, recommend they complete onboarding first by running the sample pipeline. Once onboarding is done, they will be recommended to build their own pipeline.
Assumption: By the time this skill runs, the project has been scaffolded, the user is logged in to dltHub, and the playground workspace is connected. Steps 1–2 are complete.
Anti-patterns
- ❌ Proceeding to Step 4 before Step 3 is confirmed complete — wait for
dlthub runto finish and confirm the run succeeded before opening the browser. - ❌ Treating Step 4 as optional — opening the dataset browser is the final onboarding checkpoint, not a nice-to-have.
- ❌ Running
uvx dlthub-init@latestwithout a confirmed directory — always ask and wait for the user's answer before executing. - ❌ Continuing to assist in this session after
uvx dlthub-init@latestcompletes — this session is scoped to theone-shotproject and theplaygroundworkspace and does not have the new project's toolkit, MCP servers, or skills loaded. Do not proceed. Tell the user to close this session and open a new one.
Orientation
Print this to the user before doing anything else:
- Scaffolded the example dltHub project and created a virtual environment
- Signed up / logged in to dltHub and connected to the playground workspace
- Deploy and run the sample pipeline
- Open the dltHub dataset browser
Step 3 — Deploy and run
Print to the user: Let's continue your onboarding journey.
Print to the user: - [ ] Deploy and run the sample pipeline
Deploy:
uv run dlthub deploy
Summarize which jobs were created or updated.
Run:
Print to the user: Running your pipeline and preparing your data in the background. Your data is loading into a dltHub-hosted playground destination — zero setup and no credentials needed, but it's limited to playground/onboarding use, not production.
Run both commands at the same time — start serve_headless.py in the background, then run the pipeline in the foreground:
uv run dlthub run load_sample_shop -f &
uv run .scripts/serve_headless.py jobs.onboarding_success
The -f flag streams logs in real time. Wait for the job to complete.
When the run starts, the CLI prints metadata including a run_url (a link to this run on dltHub Platform) before the log stream. After the job finishes successfully, print to the user:
You can also view this run on the platform:
Print the URL on its own line as plain text — not in backticks, not as a markdown link — so it renders as a clickable link.
If it fails:
uv run dlthub job logs load_sample_shop
| Error | Cause | Fix |
|---|---|---|
Trial period has ended |
Plan expired | Contact support@dlthub.com |
| Workspace connection error | Not connected, or connected to the wrong workspace | Run uv run dlthub workspace list to see each workspace's name and ID. If multiple workspaces are listed, pick the personal playground workspace from dlthub-start (not a shared org workspace), then connect explicitly: uv run dlthub workspace connect <workspace_id> |
Print to the user: - [x] Deploy and run the sample pipeline
Step 4 — Open the dltHub dataset browser
Once Step 3 is fully complete, print to the user: - [ ] Opening dltHub dataset browser
Retrieve the workspace ID if it is not already known and note it for reference:
uv run dlthub workspace info
Then open the dataset browser:
uv run .scripts/show_notebook.py jobs.onboarding_success
Print to the user: - [x] Opening dltHub dataset browser
Onboarding complete — what's next?
After Step 4 completes, immediately print to the user:
"Onboarding complete! Your data lives in the dltHub-hosted playground destination — when you build your own pipeline, you'll choose a real destination of your own. When you're ready to continue, ask me:
Help me get started building and running a data pipeline on dltHub"
When the user says "Help me get started building and running a data pipeline on dltHub"
Tell the user: "Amazing, let's get started!"
Then explain the following steps in order:
Run
uvx dlthub-init@latestin whatever directory they'd like their new project to live in.- Ask the user for the target directory and stop and wait for their answer before running anything.
- The command may print a prompt about creating a virtual environment after listing the created files — this is part of the normal scaffolding process. If files were listed, scaffolding succeeded; do not re-run. Only re-run if no files were listed at all.
Once scaffolding is done, tell the user: "Please close this Claude Code session now and open a new one from within the new project directory." Do not continue assisting in this session — it is scoped to the
one-shotproject and does not have the new project's toolkit, MCP servers, or skills loaded. Any further help here will be missing the agentic capabilities the new project needs.In the new session, they should tell the agent:
"Help me build and deploy a minimal pipeline"— that will guide them through creating a custom source and destination in under 5 minutes.