CtrlK
BlogDocsLog inGet started
Tessl Logo

cluster-onboarding

Interactive onboarding runbook for standing up FoliaNexa's public-facing infrastructure — a VPS edge (WireGuard tunnel, Caddy TLS, portal), its DNS records, and trusting one or more LXD hosts (e.g. 3 machines) into the cluster. Use when the operator says things like "help me set up my VPS", "configure DNS", "onboard my LXD servers", "bootstrap the cluster infra", or wants to walk through CLAUDE.md's Phase 0/3/9 or docs/vps-edge-deployment.md interactively instead of reading the docs cold. Drives real commands on whichever machine the session is running on (VPS, an LXD host, or a jump box with SSH to both), tracks progress across machines/sessions in a small local checklist file, and never invents infrastructure details (domains, IPs, tokens) — always asks.

78

Quality

98%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Passed

No findings from the security scan

SKILL.md
Quality
Evals
Security

Quality

Content

96%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

A strong operational runbook: fully executable commands with exact flags, explicit verification gates and error-recovery fallbacks at every phase, and honest handling of multi-machine session reality. The only weaknesses are minor trimmable rationale prose and an all-inline structure where a couple of sections (the labels workaround, the checklist schema) could live in reference files.

Suggestions

Trim the rationale prose — the intro paragraph justifying what the skill adds and asides like 'catching a bad --address on host 1 is much cheaper...' could be cut to one line each without losing actionable content.

Move the full checklist YAML schema (§0) and the §4.3 labels/--skip-enroll enrollment workaround into a small reference file (e.g. references/checklist-schema.md and references/host-enrollment.md), keeping SKILL.md as the sequenced overview.

The §4.3 narrative about the nonexistent `folia-nexa-mgmt hosts add` CLI subcommand is time-sensitive ('as of this writing') — isolating it in a clearly labeled reference or 'known gaps' section would keep the main workflow stable if the CLI changes.

DimensionReasoningScore

Conciseness

The body is dense with project-specific operational detail and never explains concepts Claude already knows, but the intro paragraph ("It doesn't reinvent any mechanics... rather than making the operator copy/paste everything by hand with no memory of what's already done") and scattered rationale asides (e.g. "catching a bad --address on host 1 is much cheaper than discovering it after all 3 are 'done.'") could be trimmed. This fits the efficient-with-minor-over-explanation anchor rather than the lean every-token-earns-its-place anchor.

4 / 5

Actionability

Commands are fully executable with exact flags: `apt install wireguard-tools caddy`, `deploy/vps/setup-wireguard.sh --role home --peer-public-key <...> --peer-endpoint <vps-public-ip>:51820`, `tools/folia-host-join.sh --mgmt-url ... --join-token ... --name ... --address ...`, `caddy validate --config Caddyfile --adapter caddyfile`, plus a concrete three-row DNS A-record table and a copy-paste-ready curl fallback for enrollment. Placeholders are only used for values genuinely only the operator knows.

5 / 5

Workflow Clarity

Phases are explicitly sequenced (§0–§5, 4.1–4.4) with state tracked in a checklist that is only flipped "once each is *verified*, not just attempted". Validation checkpoints are explicit throughout (`wg show` recent handshake, `curl -I https://admin.<domain>/healthz`, `dig +short`, `folia-nexa-mgmt hosts list`), with feedback loops for error recovery (enroll 401/404 → fall back to `--skip-enroll`; `lxc info` failure → run `lxd init` first) and out-of-order handling (curl verification gated on DNS being live).

5 / 5

Progressive Disclosure

Sections are well-signaled and navigable with a logical progression from checklist bootstrap to per-phase runbooks, but it is a single ~240-line inline file with no one-level-deep reference split at all — content like §4.3's labels/curl workaround or the full checklist schema are inline candidates for a separate reference. Good structure with minor organization gaps fits the score-4 anchor; it is not a monolithic reference dump, so not 3.

4 / 5

Total

18

/

20

Passed

Description

100%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

An exemplary description: it states precisely what the runbook covers, gives explicit and natural trigger phrases for when to use it, and even documents operational behavior (which machine it runs on, cross-session tracking, never inventing infra details) — all in third-person voice. No vague fluff or over-claims.

DimensionReasoningScore

Specificity

The description enumerates concrete actions: standing up a VPS edge ("WireGuard tunnel, Caddy TLS, portal"), its DNS records, and trusting LXD hosts into the cluster, plus how it operates ("Drives real commands on whichever machine the session is running on", "tracks progress across machines/sessions in a small local checklist file", "never invents infrastructure details (domains, IPs, tokens) — always asks"). Coverage across the domain is comprehensive with no generic filler.

5 / 5

Completeness

Both questions are answered explicitly: the opening sentence states exactly what the skill does, and "Use when the operator says things like..." gives concrete trigger phrases for when to invoke it. This matches the anchor for a clear what AND when with concrete triggers.

5 / 5

Trigger Term Quality

It quotes the exact natural phrases an operator would say: "help me set up my VPS", "configure DNS", "onboard my LXD servers", "bootstrap the cluster infra", plus doc-entry triggers ("CLAUDE.md's Phase 0/3/9", "docs/vps-edge-deployment.md"). Natural terms and variations are covered across all three sub-tasks, matching the comprehensive-coverage anchor.

5 / 5

Distinctiveness Conflict Risk

The skill occupies a clear niche — FoliaNexa cluster infrastructure onboarding — with project-specific triggers (named docs and phases) and quoted operator phrases. It is highly unlikely to fire for an unrelated skill.

5 / 5

Total

20

/

20

Passed

Validation

100%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository
kenvandine/FoliaNexa
Reviewed

Table of Contents

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.