Under the hood
Useful2 Signal3
A message from another session that your receive policy turns away is now reported as refused.
What's wrong with this entry?
Anonymous. No account, no email.
What
The internal SDK event stream that reports command progress gained a refused state for a peer message turned away by the session's receive-side policy before it ever joined the queue, so it is never preceded by "queued". The existing discarded state now also covers a session ending while such a message is still held.
Details
- The schema documents a known gap: the queue's own admission checks (rate limit, duplicate collapse, loop guard, queued-peer cap) can leave a message in "queued" with no terminal state.
- The git-operation event docs now note that pushing several branches emits one event per branch.
Evidence
declined by the session's receive-side policy before entering the queue
Strings lifted out of the shipped bundle, so the claim above can be checked against them.