Projects¶
A project is a folder for related work: a qualification lot, a fixture study, a customer return. Everything you plot or publish hangs off a project.
Create and open¶
In the app: open a team → New project. Give it a name the lab will recognize (cable-qual, not test).
In the SDK, after cs.set_org and cs.set_team:
project = cs.projects["cable-qual"]
Names and slugs are how you navigate. If the name is unique on the team you can use it in brackets; otherwise open the project from the team list in the app and copy the slug.
What lives in a project¶
- Datasets — measurements and published results
- Notebooks — overlays, notes, snapshots
- Members — who can see this project if it is not team-wide
A team-wide project is visible to everyone on the team. A members-only project is visible to the people you add (plus owners/admins who manage the team).
Good defaults¶
- One project per campaign, not per file.
- Put computed outputs in the same project as the source measurements so reviewers do not hunt.
- Use dataset parameters (
lot,temp,sn) for the variables that change inside the campaign. Do not explode one lot into dozens of tiny projects.
SDK convenience¶
cs.set_project("cable-qual")
# later:
cs.publish(traces=[result], name="filtered-results")
cs.set_project is the default destination for cs.publish in that session. Libraries and tests should prefer an explicit Client(team="rf-lab") so two scripts do not share that default.