Oci os management

Skill adibirzu/oci-skills/skills/oci-os-management

Operate OCI OS Management Hub for patching and update governance across Oracle Linux, Ubuntu, and Windows managed instances. Use for OS Management Hub registration, profiles, groups, dynamic sets, lifecycle environments, software sources, Ksplice, scheduled update jobs, reports, management stations, Oracle Cloud Agent plugin readiness, and patch compliance evidence.From its SKILL.md

Install

npx -y skills add adibirzu/oci-skills --skill oci-os-management

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

6 stars6 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

5.3 KB, as published. Nobody here has run it

OCI OS Management

Operate OS Management Hub as a fleet update and compliance owner. Compute instance lifecycle, VCN reachability, and image selection remain with oci-networking-compute; release/compliance decisions remain with oci-security-compliance.

Routing

Intent	Owner
OS Management Hub registration, profiles, groups, dynamic sets, software sources, lifecycle environments, update jobs	This skill
Ksplice enablement and update governance for Oracle Linux	This skill
Managed-instance reports, patch compliance, event and job troubleshooting	This skill
Compute instance launch, terminate, shape, image, VNIC, Oracle Cloud Agent plugin reachability	oci-networking-compute
Compliance threshold, exception, audit evidence, and vulnerability finding policy	oci-security-compliance
Terraform HCL and state ownership	oci-terraform-authoring

Read os-management.md before rendering an exact command or patching runbook.

Workflow

Confirm named context with ./scripts/oci_preflight.sh -c "$COMPARTMENT_OCID"; stop if the resolved tenancy/compartment does not match the intended target. Also confirm operating-system family, support eligibility, network path, update window, reboot policy, and rollback owner.
Read managed instances, profiles, groups, dynamic sets, lifecycle environments, software sources, scheduled jobs, events, reports, and agent plugin state.
Treat missing managed instances as inconclusive until registration profile, Oracle Cloud Agent or Management Agent state, permissions, network, and retention windows are checked.
Validate exact installed OCI CLI or SDK shapes before presenting commands.
Classify profile/group/source creation as additive, schedule/content changes as in-place, install keys as credential, and reboot/removal/update execution as destructive when it can disrupt workloads.
Verify job terminal state, package/security-update result, reboot outcome, health, and compliance report before closing.

Common multi-step flows

Request	Sequence
Register OCI instances	preflight -> read plugin/profile/network -> create or select profile -> register -> verify managed instance and report state
Create a patch baseline	inventory OS/support -> choose software sources -> lifecycle environment -> group/dynamic set -> schedule -> dry-run/readiness report
Run updates	maintenance window -> job preview -> approval -> update job -> terminal state -> reboot/health checks -> compliance evidence
Enable Ksplice	verify Oracle Linux support -> software sources -> group/profile policy -> staged update -> verify no-reboot patch status
Troubleshoot missing hosts	read events -> agent/plugin state -> profile/group association -> network/service gateway -> IAM -> retention window

Safety

Never print hostnames, private IPs, OCIDs, package inventory that identifies a customer, management-station endpoints, install keys, or support identifiers.
Update jobs can reboot or destabilize workloads; require a maintenance window, owner approval, rollback plan, and post-update health checks.
Every mutation runs through run_action --risk <additive|in-place|credential|destructive> --compartment <COMPARTMENT_OCID> --description "<...>" -- oci_cli ... (honors OCI_SKILLS_DRY_RUN=true for a no-op preview); an update job that can reboot or disrupt workloads additionally requires explicit confirm.
MACS install keys and comparable registration secrets are credential risk and must use 0600 files or Vault-backed handling.
Do not use OS Management Hub against unsupported platforms as proof of patch compliance. State coverage limitations explicitly.
Terraform remains the durable owner for profiles, groups, schedules, and related infrastructure where Terraform is declared.

Verification and rollback

Verification includes managed-instance lifecycle, agent heartbeat, job terminal state, package/security-update report, reboot status, service health, and compliance evidence. Rollback means restoring the previous lifecycle/profile or software-source selection and using OS/vendor rollback procedures where available; never promise package rollback without tested owner evidence.

Official documentation

OS Management Hub · Overview · Getting started. Full list in the OS Management reference.

Open Knowledge Format grounding - every doc link here is registered and liveness-checked in the oracle-docs.md index (the pack's single source of truth). When extending this skill, cite the most specific official page through that index; the non-official MCP gateway is never a source of truth.