What problem does it solve? Cloud agent containers can be reclaimed at any moment, wiping local state and killing long-running coordination work. This Skill turns a single Claude Code cloud session into a durable coordinator that survives container resets by keeping every piece of state — cursors, duty scripts, grants, checkpoints — on the Fulcra bus instead of the machine. ## Core Features & Use Cases - Container-reset survival: A stash-and-bootstrap pattern restores duty scripts from the Fulcra store after any reset, with a fail-closed recovery ritual embedded in every standing prompt. - Durable wake scheduling: Three-layer wake design (server-side Routines, self-chained one-shots, session cron) ensures standing duties like watchdogs and blocked-work sweeps never silently die. - Bus-native identity and queue discipline: Presence beats, cursor-based queue reads with staged commit tokens, and durable-first dispatch keep obligations tracked across sessions. - Use Case: An operator needs a fleet coordinator that wakes hourly, dispatches and reviews work, and keeps running even after its cloud container is destroyed mid-task — this Skill provides the full setup, recovery, and verification doctrine. ## Quick Start Set up my Claude Code cloud session as a persistent Fulcra bus coordinator that survives container resets and wakes on schedules.
