webnovel-resume_skill

Overview

This skill recovers interrupted webnovel tasks with precise workflow state tracking and guided user confirmation. It detects where a long-form writing workflow stopped, presents safe recovery options, and executes the chosen recovery without guessing or auto-resuming. The goal is to minimize data loss while avoiding risky automatic continuation.

How this skill works

The skill first verifies it is operating from the project root and that the workflow state file exists. It runs a detection routine to identify the interrupted step and collects metadata (command, args, timestamps, completed/remaining steps). It then shows the user curated recovery options with risk levels and executes the selected safe recovery commands only after explicit confirmation.

When to use it

• A writing or build task stopped unexpectedly and you want to resume safely.
• You find partial chapter files or inconsistent state after a crash.
• You need to decide between deleting partial output or rolling back to a stable point.
• You want a guided, auditable recovery procedure rather than blind auto-resume.
• Long downtime occurred (>1 hour) and you need to assess context loss risk.

Best practices

• Always run the skill from the project root where the workflow state file is present.
• Inspect the presented interruption report before choosing any recovery action.
• Prefer the recommended 'delete and restart' option when context loss risk is high.
• Avoid automatic or AI-driven continuation of partial content; require explicit user consent.
• When in doubt, choose rollback to the last stable checkpoint rather than in-place edits.

Example use cases

• A chapter generation halted mid-write; view interruption details and choose cleanup or rollback.
• State file shows partial updates after a crash; run detection, repair or revert safely.
• A multi-step review process was interrupted; decide to re-run the review or skip and continue.
• Long interruption (>1 hour) where context may be stale; follow the recommended delete-and-restart flow.
• High-cost recovery step failed; present cost-aware options so the user can choose cheaper or safer paths.