Recover a stack
llms.txtlists every page for an agent
A run crashed
Section titled “A run crashed”Re-run it. Its pending operations stay on the resources it touched and block changes to their dependents until cleared (Pending operations and crashes).
A machine is lost
Section titled “A machine is lost”The records live in the store, so log in to the same plunge+ URL elsewhere.
The exception is plunge+file://, which lives on that one host.
A record was overwritten or deleted
Section titled “A record was overwritten or deleted”Object versioning, which the store requires unless waived, keeps each
record’s earlier versions. Restore one with the cloud’s own tools, then run
plunge preview; there is no plunge stack history
(Records).
The whole store is lost
Section titled “The whole store is lost”If a stock checkpoint of the stack exists, adopt it again (Adopt an existing stack).
Optional: hand this page to your coding agentThe steps work by hand too.
Follow https://intentius.io/plunge/use/recover/ for the failure I describe and tell me which section applies. Change nothing.
Never run `plunge up`, `plunge destroy` or `plunge stack adopt --yes`, approve or apply a saved plan, or merge; never edit or remove a plunge-stack or plunge-urn tag or label by hand.