backup-disaster-recovery-and-data-migration
Skill8 starsChanged 5 months agoMITView on GitHub
Plan, implement, test, and audit backups, disaster recovery, VPS or cloud migrations, restore drills, RPO/RTO targets, data consistency, and retirement of old resources. Use before any recovery, replacement, resize, cross-cloud move, backup policy change, or destructive cleanup.
In this repository68 other files it gives its agents.
- animejsSkill
- sketch-animation-videoSkill
- sound-fx-for-videoSkill
- content_series_infographicSkill
- tech-article-seriesSkill
- ansible-playbookSkill
- gin-goSkill
- github-actionsSkill
- iac-hclSkill
- nodejs-nextjs-typescriptSkill
- postgresql-developmentSkill
- programming-paradigmsSkill
- python-shellSkill
- agent-team-harnessSkill
- ai-agent-collaboration-standardSkill
- ci-cd-workflow-specSkill
- config-as-code-specSkill
- execution-ownership-migrationSkill
- feature-lane-release-workflowSkill
- generation-upgrade-acceptanceSkill
- gitops-canonical-resource-deliverySkill
- harness-workflowSkill
- infrastructure-as-code-specSkill
- issue-pr-traceability-standardSkill
- multi-environment-delivery-and-releaseSkill
- project-development-standardSkill
- store-and-startup-homepage-specSkill
- zero-trust-overlay-deliverySkill
- it-infra-continuous-pngSkill
- image-svg-pptx-proSkill
- autoglm-websearchSkill
- blogwatcherSkill
- Browser AutomationSkill
- docxSkill
- find-skillsSkill
- image-cogSkill
- image-resizerSkill
- pdfSkill
- pptxSkill
- Self-Improving Agent (With Self-Reflection)Skill
- skill-security-checkSkill
- skill-vetterSkill
- video-translatorSkill
- wan-image-video-gen-editSkill
- xlsxSkill
- capacity-cost-and-resource-lifecycleSkill
- github-actions-operational-dispatchSkill
- incident-response-and-change-managementSkill
- network-dns-tls-edge-managementSkill
- observability-slo-and-alertingSkill
- secrets-identity-and-access-governanceSkill
- service-catalog-and-runbook-standardSkill
- system-db-three-role-securitySkill
- unified-observability-architectureSkill
- vault-eso-tls-certificate-governanceSkill
- ai-tech-news-videoSkill
- ai-ti-wo-shuo-explainerSkill
- it-infra-evolution-videoSkill
- it-infra-evolution-video-v2Skill
- product-intro-videoSkill
- The first 60 shown.
The file
We split this skill into the rules it gives an agent
5 rules. About 0 are common, 5 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈5
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 24
Verify backup completion, encryption, retention, immutability/access separation, and restore readability.
- Distinctive: line 25
Never store backup credentials, unseal keys, or database passwords in Git, CI secrets, artifacts, or local temporary files.
- Distinctive: line 30
Run scheduled restores into an isolated target.
- Distinctive: line 37
Create and verify the replacement.
- Distinctive: line 50
Keep the source service, schema, and writer unchanged during candidate acceptance.
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Backup, Disaster Recovery, and Data Migration
A backup is not a recovery capability until a restore has met its declared RPO, RTO, integrity, and security requirements. Keep backup, migration, cutover, and cleanup as independently auditable phases.
For generation replacement, read Generation upgrade operational validation for target protection, populated-data restore, actual incident regression cases, writer handoff, and post-cutover reconciliation. The linked engineering skill owns stage ordering and acceptance gates; this skill owns recovery procedures.
Reference basis: NIST contingency planning guidance for coordinated recovery of systems, operations, and data after disruption.
1. Establish the recovery contract
For every critical service, define owner, data classes, dependency order, RPO, RTO, backup frequency, retention, encryption/key owner, restore destination, validation query, and last successful drill. Use the business impact to prioritize recovery; do not use a single generic target for all services.
2. Back up for restoration
- Capture application-consistent data: database logical/physical backup, object data, configuration, identity/Vault recovery material, and artifact/digest references as appropriate.
- Verify backup completion, encryption, retention, immutability/access separation, and restore readability. A successful upload alone is not verification.
- Never store backup credentials, unseal keys, or database passwords in Git, CI secrets, artifacts, or local temporary files. Use OIDC → Vault short-lived access.
- Track backup provenance: source instance ID, volume/data size, application version, timestamp, checksum or integrity result, and restore compatibility.
3. Drill restores
Run scheduled restores into an isolated target. Measure actual RPO/RTO, validate schema/application health and representative data, then destroy the drill target only after recording results. Treat a failed or overdue drill as a recovery risk, not an administrative task.
4. Migration and resize decision
- Preflight source/target capacity, architecture, OS/runtime compatibility, required ports, DNS, and dependency order.
- Classify upgrades using the engineering UI/API/DB release rules. Database engine versions, SQL schema transitions, and upgrade data migrations require a new cluster, independent storage, and an empty target database. Vault Raft and other persistent-volume workload version/state transitions use the same replacement boundary. A backup or provider-supported operation does not authorize an in-place stateful business upgrade. UI/API code-only fixes and the explicitly independent non-business/non-billing monitoring service may follow their allowed in-place release path.
- For downgrades, cross-cloud moves, or target disks smaller than the source, use application-level backup and restore. Do not assume a volume snapshot can restore to a smaller disk.
- Create and verify the replacement; adopt it into Terraform state; regenerate CMDB/inventory; deploy configuration; run data and service health checks.
- Cut traffic/DNS only through an explicit, reversible decision. Retain the source during an observation window, then enumerate and explicitly approve its deletion.
5. Recovery execution
Use a written runbook with preconditions, exact backup identifier, target scope, ordered dependencies, validation, rollback/abort point, and communication owner. Do not overwrite a healthy source or clear initialization state merely because a key file or local marker is missing; stop and make a recovery decision with evidence.
Record every migration or restore with source/target IDs, chosen backup, planned and actual RPO/RTO, validation evidence, DNS/cutover time, and old-resource cleanup approval.
6. Stateful upgrade execution and recovery
- Before migration, verify a protected source backup restores into isolation; retain a recoverable target baseline before modifying an existing candidate. After data copy, back up and restore a populated target and verify representative data, schema, permissions, and application connectivity. An empty-database restore alone does not establish business recovery.
- Initialize the reviewed SQL schema only on the new empty target, then use the business owner's fixed, validated migration tool (for example,
migratectlwhere that is the existing owner tool). Pin its scope, version, schema checksums, source snapshot boundary, and target identity. Do not install or assume a particular CLI for unrelated systems. Keep target independent writers paused during transfer as required by its actual contract. - Keep the source service, schema, and writer unchanged during candidate acceptance; read-only migration credentials do not freeze source business writes. Final catch-up, source-writer fencing, target-writer handoff, and traffic cutover require a separately reviewed operation within explicit authorization.
- Retry according to the actual tool contract. An empty-target-only copy is not an idempotent import into a populated database. After an uncertain result, inspect receipts and data first; never truncate/reset the target to make a retry pass. Prove rollback and emptiness before retrying an empty-target copy.
- Code/image rollback and SQL down migration do not prove data recovery. Preserve failed targets and evidence; recover into a separate isolated target from a verified backup or use an explicitly reviewed reconciliation procedure. Once a target accepts new writes, the old source is behind; switching traffic back does not restore those new facts.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

