Rating
Visible 0–10 self-rating
Not a certification score. It shows how comfortable I feel solving real problems in that area.
Self-assessed · intentionally honest
What I know, where I still need help, and what I would trust myself to change.
The numbers are my own assessment. More important is the context beside them: how independently I work, where I use documentation or AI, and how cautious I would be before touching a production system.
How I use AI: mostly as a speed multiplier, second opinion, or a way to find reasoning gaps faster. Where I depend on it heavily, I say so. If I cannot explain and verify a change, it belongs in the lab first.
Independence
What I actually rely on
Memory, documentation, search or AI. Generated output only counts after I understand and verify it.
Production
How cautious I would be
Independent, review-first, sandbox-first or lab-only depending on the technology.
Infrastructure & operations
Skill
Self-rating
How I work
Production posture
Evidence
Linux & shellServices, users, permissions, logs, packages, systemd, troubleshooting.
Independent day to day; documentation for uncommon flags or tools I have not used in a while.
Independent for familiar changes; verify before and after.
NetworkingRouting, NAT, DNS, firewalling, private access and VPN paths.
Very comfortable with the networks I actually run; docs/AI mainly for edge cases and design checks.
Independent on familiar topologies; review-first for more complex VLAN/switching designs.
Docker / ComposeCompose files, logs, mounts, networking, health checks and recovery.
I can write, edit and reason about Compose files myself; docs/search when useful.
Independent for normal service changes.
Operational configs are partly private by design.
Kubernetes / k3sDaily running private cluster, kubectl, manifests, services, ingress and troubleshooting.
Active hands-on lab use; setup and many changes are still heavily AI-assisted.
Lab/sandbox first. Changes are verified against the real lab cluster rather than accepted blindly.
Proxmox VEVM lifecycle, templates, networking, lab systems and virtualization concepts.
Strong virtualization understanding; I can operate Proxmox without AI while product-specific depth is still growing.
Independent for normal VM/lab work; review complex storage/network changes.
Terraform / OpenTofuUsed for infrastructure definitions in my lab.
I work with it while AI currently guides most authoring. I inspect the configuration and plan, run it, and validate the result rather than treating generated code as independent skill.
Lab only / sandbox-first.
AnsibleInventories, roles, bootstrap and configuration concepts used in lab work.
I have worked with and use Ansible, but AI still guides most implementation. I review what it produces and validate the run instead of claiming independent authoring.
Sandbox-first; no blind production runs.
SMB / NASShares, permissions, migrations and practical storage administration.
Mostly independent; little AI needed for normal administration.
Independent with normal change/backup discipline.
Private operational infrastructure.
NFSMounts, k3s storage paths, backups and service integration.
Newer area; comfortable with what I run, while unfamiliar behavior still sends me to docs/search.
Review-first outside known patterns.
Used actively in private infrastructure.
Prometheus / GrafanaBasic setup, dashboards, exporters, probes and troubleshooting.
I can set up the basics manually; for deeper tuning I am still learning and experimenting with AI assistance.
Review-first / sandbox-first for larger changes.
NetBird / VPN / private networkingPeers, routes, access rules and private administration paths.
I can build routing and rules myself; AI mainly accelerates reasoning or catches logic holes.
Independent for familiar deployments.
Delivery & tooling
Skill
Self-rating
How I work
Production posture
Evidence
Git / GitHub ActionsRepos, diffs, CI checks, branch/PR flow and deployment workflows.
Comfortable with normal workflows; docs/search for less common Git and Actions behavior.
Review-first for workflow/deployment changes.
Bash / scriptingOperational helpers, pipelines, diagnostics and glue scripts.
Similar to Linux: strong daily use; docs for uncommon flags or syntax I have not touched in a while.
Independent for familiar scripts; always review anything destructive.