Recovery
How to inspect and continue failed or stopped Roark attempts.
Last updated
#Continue an attempt
roark continue 123 --repo owner/repo --attempt 1If --attempt is omitted, Roark uses the latest attempt recorded in:
.roark/runs/issue/<number>/attempts.json#What continue does
continue reuses valid existing artifacts, regenerates missing or malformed phase outputs, rewrites readiness, reruns verification, and publishes only if readiness and verification both pass.
If an attempt stopped as failed-verification while fix budget remains, continue plans the next fix pass, final review, readiness, and publish gate instead of just rerunning the failed command.
#Dirty workspaces
A failed attempt may leave uncommitted edits in the managed workspace. continue is allowed to recover that selected failed attempt. A fresh auto run refuses dirty existing issue workspaces and tells you to use continue or clean/remove the workspace.
#What cannot be recovered
Uncommitted failed work is recoverable only while the persistent issue workspace still exists. If the workspace is deleted, Roark can recreate it from the local or remote issue branch, but deleted uncommitted edits are gone.
#Useful files
.roark/runs/issue/<n>/attempts/<k>/attempt.json.roark/runs/issue/<n>/attempts/<k>/summary.json.roark/runs/issue/<n>/attempts/<k>/verification.md.roark/runs/issue/<n>/attempts/<k>/verification-before-fix-<pass>.md.roark/runs/issue/<n>/attempts/<k>/readiness.md
See Artifacts for the full layout.
#Recovery checklist
- Open
summary.json. - Open
readiness.mdandverification.md. - Inspect the managed workspace if uncommitted edits matter.
- Fix host setup, config, hook, ignored-file, or code issues.
- Run
roark continue.
#Next steps
- Use Troubleshooting for symptom-specific recovery.
- Use Managed workspaces before deleting workspaces.
- Use Operations runbook for scheduled-run failure response.