Skip to content

Deploy runs write to the team continuum, so an app must know a run execution origin to find its own runs #2108

Description

@joshuadarron

Deploy runs write to the team's deploy continuum (open_event_stream(..., team_id=...), log keyed {project}.{source}.{runKind}). An app that reads its runs without a team_id reads the dev stream only — so every scheduled run would be missing from the run record, which for Rocket CRM is the surface the whole product rests on. The failure mode is silent absence, not an error.

Filed from section 5.3 of the Rocket CRM runtime findings (docRevision 252). Part of the Rocket CRM build tracked in rocketride-ai/rocketride-saas#456. Named but explicitly not filed in #2045's "Related, not filed" list.

Mechanism

Where a run's events land depends on how the run was started. An app therefore has to know a run's execution origin to find it — and there is nothing in the run record itself that says an origin exists to be missed.

What we do today

Our side of this is small: pass team_id and sync both streams. We have not shipped it because schedules are blocked on the deploy env merge (#2045) anyway.

Ask

  1. A read that spans both continua for a project, or a wildcard on team_id.
  2. Failing that, document the keying, because the failure mode is silent absence rather than an error.

The platform-side question is why an app has to know an execution origin to find its own runs at all.

Related

Metadata

Metadata

Assignees

No one assigned

    Labels

    bugSomething isn't workingmodule:aiAI/ML modulespriority:mediumFeature impaired, viable workaround exists, or accepted roadmap work

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions