Labor0 Docs

Troubleshooting

Use the visible Labor0 message to fix one problem at a time without exposing private workspace data.

Safe recovery loop

  1. Confirm the selected workspace and project.
  2. Read the visible status and next action.
  3. Correct one prerequisite, permission, connection, or input.
  4. Retry only through the offered Labor0 action.
  5. If it still fails, record the page, visible status, scope name, and approximate time.

Common problems

What you seeSafe recovery
Repository is missingRepair GitHub access or register and bind the repository
Task is waitingOpen task activity and answer the visible approval or input request
Connection is unavailableReconnect it or restart indexing from Labor0
Billing total differsConfirm workspace and reporting period, then ask the billing owner
Local runner is unavailableInstall the complete Labor0 CLI bundle, then use Set up and run or Manage and troubleshoot for CLI-managed runner recovery. In the browser, Auto prefers an eligible active local runner but may use hosted fallback. An active online CLI-managed runner does not currently enable Local only because Labor0 cannot confirm readiness for the selected coding-agent runtime. Local only never falls back to hosted execution. Confirm final Local runner attribution; if hosted fallback is unacceptable, do not submit the browser task with Auto, because hosted work may begin before final attribution is available; stop before retrying.
A commit, push, or pull request appeared after the prompt said not to publishA repository-changing task follows its configured publication outcome; prompt wording can constrain content but cannot disable required publication. Do not merge unwanted work. Open Labor0 at the app, then select Projects > selected project > Graph. If the unwanted task has completed, change Status from Active work to Completed history or All statuses so the task row and its available actions are visible. If the unwanted task or any downstream task has published a branch or pull request, inspect and record every affected link before choosing Remove. If the unwanted task is still active, record those links only, open Remove, and immediately choose Remove all downstream tasks or Unblock downstream as if this task succeeded; that downstream choice cancels the active task. Before touching repository outputs, confirm that execution has stopped. If removal reports an error or you cannot confirm that execution stopped, wait for reconciliation or escalate the error before handling or backing out the recorded outputs; the Remove control may no longer be available. If changes already merged, closing the task, branch, or pull request cannot remove them from the base branch; follow repository policy to revert or otherwise back out the merged changes. Unblock downstream as if this task succeeded may immediately start newly ready successor tasks, including another publication, so use it only when that execution is intended. Remove requires Workspace Operator or above; if you are a Viewer, contact a project or workspace Operator or Admin.
Old Assistant or workspace Graph Chat link failsThe /assistant and former workspace Graph Chat routes have no redirect. Upgrade an old Tauri client manually and use Workspace Chat

Send a safe report

After the recovery loop fails, send a short report such as:

Page: Project > Tasks > task detail
Visible status: Waiting for approval
Expected: Plan review actions
Observed: Plan review unavailable after one refresh
Scope: Public docs project
Time: 2026-08-09 18:30 KST

Include only customer-visible context. Do not send credentials, tokens, private URLs or paths, prompts, transcripts, provider responses, card details, or an unreviewed diagnostic archive.

Expected result

A successful recovery changes the named prerequisite or lets the original workflow continue. A useful escalation gives the responsible owner enough visible context to locate protected details without receiving private content.

If recovery is still unsuccessful, stop repeated retries and leave the protected state unchanged. Contact support@labor0.com for best-effort help. For a reproducible download, checksum, or archive-content defect in a published CLI release, use ZephyrCloudIO/labor0-releases issues. These channels have no response-time or availability SLO.