OpenClaw Backup Skill

Use the bundled script scripts/openclaw-backup.sh for real operations. Do not rewrite backup logic inline when the bundled script already covers the task.

When to use

Use this skill when the user wants to:

• - create a local OpenClaw backup;
• create a manual backup before risky changes;
• prune old regular backups;
• schedule recurring backup runs;
• inspect or adjust retention behavior.

Preconditions

This skill assumes:

• - Linux;
• Bash 4+;
• OpenClaw CLI installed and working;
• local filesystem access.

Primary commands

• - Show the current CLI contract directly from the script:

• - Regular backup:

• - Manual backup:

• - Dry-run prune:

• - Automatic backup + prune:

Recommended scheduled run

For recurring local backups, a good default pattern is:

• - INLINECODE6

On Linux hosts, use the system scheduler (for example cron) when the goal is a direct local script run without chat delivery.

Operational notes

The current backup format includes:

• - workspace contents, including assistant identity and continuity files (for example SOUL.md, USER.md, AGENTS.md, and other workspace-level context files);
• OpenClaw config;
• cron export and human-readable cron summary;
• software versions;
• restore instructions;
• manifest.

The current backup format intentionally excludes:

• - backup storage inside a workspace (legacy support);
• temp artifacts;
• environment variables / system env / system override.

• - Regular archives are subject to prune; manual archives are excluded from normal prune.
• Retention is based on UTC timestamps encoded in archive filenames, not on mtime / ctime.
• The script resolves the OpenClaw workspace/config automatically unless the caller passes an explicit output directory.
• If the user wants recurring backups, ask before installing or changing scheduled jobs.
• For the full behavior contract, retention model, and archive naming rules, read references/spec.md.