Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Manage use when you need to work with backup and recovery. This skill provides backup automation and disaster recovery with comprehensive guidance and automation. Trigger with phrases like "create backups", "automate backups", or "implement disaster recovery".
.claude/skills/jeremylongshore-fairdb-backup-manager/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 22% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 113% | 0% |
| case-04 | ✗→✓ | ▲ Improved | 32% | 0% |
| case-11 | ✗→✓ | ▲ Improved | 36% | 0% |
| case-12 | ✗→✓ | ▲ Improved | 71% | 0% |
Automate backup and recovery operations for FairDB database instances. Generate backup scripts, configure retention policies, schedule automated backups to local storage or S3, and produce tested restore procedures with integrity verification.
tar and rsync installed for file-level backupss3:PutObject and s3:GetObject permissions (if using S3 as backup target)tar czf), and checksum generationaws s3 cp --sse aws:kms) for off-site copies/var/log/fairdb-backup.log| Error | Cause | Solution | |-------|-------|---------| | Backup lock acquisition failed | Another backup or maintenance process is running | Check for stale lock files; implement timeout-based lock with flock | | tar: Cannot open: No space left on device | Local backup destination full | Run retention cleanup; check disk usage with df -h; increase volume size | | aws s3 cp: upload failed | Network issue or expired AWS credentials | Retry with --retry 3; refresh credentials; check S3 bucket permissions | | Restore failed: checksum mismatch | Backup file corrupted during transfer or storage | Re-download from S3; verify S3 object integrity; use a different backup copy | | Database inconsistent after restore | Backup taken during active write without lock | Ensure backup script acquires a consistent snapshot lock before export |
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 16,308 | 13,226 | -19% | 1 | 1 | 0% | 2,980 | 3,643 | +22% | 0 | 0 | — |
case-02 | fail→pass | 24,451 | 34,758 | +42% | 1 | 1 | 0% | 4,112 | 8,744 | +113% | 0 | 0 | — |
case-03 | pass→pass | 22,654 | 16,323 | -28% | 1 | 1 | 0% | 2,919 | 4,275 | +46% | 0 | 0 | — |
case-04 | fail→pass | 11,052 | 7,817 | -29% | 1 | 1 | 0% | 1,681 | 2,225 | +32% | 0 | 0 | — |
case-05 | pass→pass | 10,398 | 11,014 | +6% | 1 | 1 | 0% | 1,812 | 2,371 | +31% | 0 | 0 | — |
case-06 | pass→pass | 13,401 | 15,499 | +16% | 1 | 1 | 0% | 2,331 | 3,005 | +29% | 0 | 0 | — |
case-07 | pass→pass | 16,835 | 12,884 | -23% | 1 | 1 | 0% | 2,628 | 3,287 | +25% | 0 | 0 | — |
case-08 | pass→pass | 37,046 | 14,150 | -62% | 1 | 1 | 0% | 2,333 | 2,920 | +25% | 0 | 0 | — |
case-09 | pass→pass | 13,528 | 14,589 | +8% | 1 | 1 | 0% | 2,144 | 2,846 | +33% | 0 | 0 | — |
case-10 | fail→fail | 16,778 | 11,940 | -29% | 1 | 1 | 0% | 2,442 | 2,858 | +17% | 0 | 0 | — |
case-11 | fail→pass | 9,249 | 5,711 | -38% | 1 | 1 | 0% | 1,255 | 1,705 | +36% | 0 | 0 | — |
case-12 | fail→pass | 12,669 | 20,361 | +61% | 1 | 1 | 0% | 1,979 | 3,376 | +71% | 0 | 0 | — |
case-13 | pass→pass | 11,715 | 7,083 | -40% | 1 | 1 | 0% | 1,948 | 1,718 | -12% | 0 | 0 | — |
case-14 | pass→pass | 9,827 | 10,028 | +2% | 1 | 1 | 0% | 1,794 | 2,677 | +49% | 0 | 0 | — |
case-15 | fail→fail | 35,709 | 13,317 | -63% | 1 | 1 | 0% | 3,079 | 3,199 | +4% | 0 | 0 | — |
case-16 | fail→fail | 16,966 | 10,428 | -39% | 1 | 1 | 0% | 2,191 | 2,212 | +1% | 0 | 0 | — |
case-17 | pass→pass | 8,957 | 6,062 | -32% | 1 | 1 | 0% | 1,340 | 1,784 | +33% | 0 | 0 | — |
case-18 | fail→pass | 11,152 | 5,986 | -46% | 1 | 1 | 0% | 1,473 | 1,347 | -9% | 0 | 0 | — |
case-19 | pass→pass | 17,219 | 3,606 | -79% | 1 | 1 | 0% | 2,416 | 1,409 | -42% | 0 | 0 | — |
case-20 | pass→pass | 14,037 | 13,678 | -3% | 1 | 1 | 0% | 2,246 | 3,022 | +35% | 0 | 0 | — |
case-21 | pass→pass | 17,016 | 26,084 | +53% | 1 | 1 | 0% | 3,233 | 4,881 | +51% | 0 | 0 | — |
case-22 | pass→pass | 16,576 | 11,917 | -28% | 1 | 1 | 0% | 2,182 | 2,484 | +14% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +27 percentage points is the difference between those two pass rates over the 22 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.