You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The CLI has logic to validate the structure of a workflow.
The idea here is that if a user mis-types a key like "adapatator", the CLI will catch the error and flag them.
What with all the focus on sync and portability, it's way more likely that a workflow will be machine generated now, and so won't have those typos.
But if the app adds a new key to the workflow, the CLI will refuse to execute it - even if it's a step it doesn't care about.
I would like to think about the following options:
Use the common portability spec schema to validate against. The app does have a JSON schema. If the workflow or owning project has aversion, use that, else use latest. This relates to Better coupling of portability spec with kit lightning#4734
Add a --validate flag which will validate a workflow for you, if you want it (a bit like a circle CI yaml linter app)
only validate keys that the CLI cares about, and ignore the rest
The CLI has logic to validate the structure of a workflow.
The idea here is that if a user mis-types a key like "adapatator", the CLI will catch the error and flag them.
What with all the focus on sync and portability, it's way more likely that a workflow will be machine generated now, and so won't have those typos.
But if the app adds a new key to the workflow, the CLI will refuse to execute it - even if it's a step it doesn't care about.
I would like to think about the following options:
--validateflag which will validate a workflow for you, if you want it (a bit like a circle CI yaml linter app)