The next steps after a run finishes, queued while the context is fresh
Follow-on
Part of Chief of Staff
When a managed run or an interactive session ends, a Sonnet call drafts the second-order next steps as a brief that the queue runs within minutes.
What it does
When a run finishes, it reads the run's own progress notes and queues the next steps as a brief. The model only drafts the brief. The queue then launches each one as a managed run.
Limits
It stops at depth one, allows ten a day and drops the rest rather than deferring them. The first live poll marks everything already finished as seen, so switching it on never replays the backlog. It stays off pending a redesign build.
Stack
- Sonnet call
How these figures are sourced
The outcome figure is agent-reported: it is taken from a test log or report written by the agents that built and ran the system. The status was read from the Life Tracker session ledger, 13 October decision. Every figure and claim on this page is tied to a source file or a CV line in the site's content.