tegwick
5517d91bf5
State Hub was unreachable during the offline session that bootstrapped Railiance01. Inbox event drained and T03/T04/T05 task statuses synced to the hub (C-10 drift fixed via check_repo_consistency --fix). Progress event id: de18d727-eea5-4dfa-913c-8fe62245cda4 Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
state-hub-inbox
Offline event queue for State Hub sessions run without an active reverse tunnel.
When the State Hub MCP server is unreachable, Claude writes pending events here
as YAML files instead of dropping them. On the next connected session, these are
drained by calling the appropriate State Hub write tools and marking each file
status: ingested.
File format
type: progress_event # or: decision, task_status_update
topic_id: <uuid>
workstream_id: <uuid>
event_type: milestone
summary: "..."
detail: {}
status: pending # → ingested after processing
recorded_at: "YYYY-MM-DD"
source_repo: railiance-hosts
Drain procedure (manual until automated)
At session start, if get_domain_summary() succeeds, check for pending files:
grep -l "status: pending" state-hub-inbox/*.yaml
For each, call add_progress_event() with the file's fields, then update
status: pending → status: ingested and commit.
See also: contrib/feature-requests/fr-2026-03-08--threephoenix--state-hub--offline-inbox-ingest.md