levnikolaevich/claude-code-skills / plugins/delivery-suite/skills/ln-63-deployment-engineer
ln-63-deployment-engineer
VerifiedPrepares delivery pipelines and remote infrastructure, then executes authorized rollouts; not standalone quality-gate setup or local container packaging.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 11.3 KB |
| references/manual-rollout.md | markdown | 4.1 KB |
references/manual-rollout.md
View on GitHub# Manual rollout
Use when the user requests deployment through direct server access or an existing provider CLI/API without the normal pipeline. Consume the main skill's authorization, evidence and recovery contracts. Do not introduce CI, publish a repository or trigger paid jobs merely to deploy manually.
## Build and transfer
- Bind local checks and the artifact to one source revision. Export that revision to an isolated build context when the checkout contains unrelated changes; a successful build of a dirty checkout does not prove committed provenance.
- Inspect the build definition and CI for version arguments and defaults. When building manually, supply the intended source-derived application version and revision metadata through the existing build mechanism; verify their values in the resulting artifact. If no reliable mapping is available, record the gap instead of treating a default or date-like string as proof of the revision.
- Choose the build location using actual CPU, memory, disk, architecture and build tooling. Prefer local or existing build-host packaging when a production build would compete with the running service. Reuse the project's build definition and compatible target platform.
- For container delivery without a registry, save and transfer the image through authenticated SSH/SFTP or an equivalent approved channel. Verify the archive checksum and loaded artifact identity before replacement. Account for platform-specific image config digests versus multi-platform index digests; compare the same identity, not merely a mutable tag. Preserve a revision label or equivalent build metadata.
- Inspect permissions and free space before transfer. Use a private writable staging directory. Keep secrets out of arguments, logs and source archives; configuration backups containing secrets need restricted access. Remove only run-owned staging files after verification.
## Switch and recover
- Record the running artifact, deployment configuration and automatic updater state. Preserve a usable rollback artifact and configuration before replacing either. Data migration compatibility follows the project's existing policy; a container rollback does not undo database changes.
- Prevent automatic reconcilers or image updaters from replacing the manual release with a stale remote artifact. Use the platform's supported pause, pin or exclusion mechanism, scoped to affected services. A running updater is safe only when its effective configuration cannot undo the manual release.
- Change only the intended artifact references and necessary updater settings. Preserve runtime secrets, mounts, networks, resource limits and logging. Validate the effective configuration without printing sensitive values.
- Roll out coupled API, worker and scheduler processes according to their migration and compatibility contract. Pin them to the intended immutable artifact. Avoid an implicit pull of a mutable remote tag during recreation.
- If health fails, stop forward rollout and use the preserved artifact through the evidenced recovery path. Keep updater protection active during rollback; restoring an old configuration that references `latest` can otherwise replace the rollback again. Verify the recovered service, not just the rollback command.
## Prove the outcome
- Verify actual running artifact identity, applicable readiness checks, relevant startup errors and required smoke behavior. Check the exposed application version when available, using the main skill's identity-proof rule. Observe restarts and updater behavior for the project's required window. A successful upload or container start alone does not prove a healthy deployment; report completed external mutations separately from unverified health.
- Reconcile the project runbook with the effective manual mode and record how a later authorized release resumes the pipeline or updater safely. Do not re-enable an updater until its target artifact is verified as compatible with the intended release.
- Report deployment separately from commit, push, registry publication and local packaging. Keep provider addresses, service names, concrete commands and credentials in their project-owned sources, not this general skill.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| writing-skillsobra/superpowersUse when creating new skills, editing existing skills, or verifying skills work before deployment | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| xlsxanthropics/skillsUse this skill any time a spreadsheet file is the primary input or output. This means any task where the user wants to: open, read, edit, or fix an existing .xlsx, .xlsm, .xltx, .csv, or .tsv file (e.g., adding columns, computing formulas, formatting, charting, cleaning messy data); create a new spreadsheet from scratch or from other data sources; or convert between tabular file formats. Trigger especially when the user references a spreadsheet file by name or path — even casually (like "the xlsx in my downloads") — and wants something done to it or produced from it. Also trigger for cleaning or restructuring messy tabular data files (malformed rows, misplaced headers, junk data) into proper spreadsheets. The deliverable must be a spreadsheet file. Do NOT trigger when the primary deliverable is a Word document, HTML report, standalone Python script, database pipeline, or Google Sheets API integration, even if tabular data is involved. | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| shipping-and-launchaddyosmani/agent-skillsPrepares production launches. Use when preparing to deploy to production, or when asking what needs to be in place before shipping. Use when you need a pre-launch checklist, when setting up monitoring, when planning a staged rollout, or when you need a rollback strategy. | Stars: 119k | Updated | Claude CodeCodex | Verified | Install |
| childhoodthedotmack/claude-memBabysit a shipped release after merge, publication, or deployment. Verify published packages, production behavior, installed plugins, and late CI or user reports until the observation window is clean; use babysit for PR readiness before release. | Stars: 118k | Updated | Claude CodeCodex | Verified | Install |