Claude Skills Inventory

Cosa Loris e team trovano installato il giorno 1 quando aprono Claude Code in questa cartella o in D:\Desktop\Multiverbe\. Se introduci una skill nuova, aggiungi una riga qui + log in execution/decisions-log.md.

Come Loris (o chiunque) usa una skill · 3 step

Apri Claude Code in una cartella Multiverbe
Scrivi una richiesta in linguaggio naturale (es. "scrivi cold email a un parrucchiere Lugano")
Claude trigghera automaticamente la skill giusta in base alle keyword

Non devi memorizzare i nomi. Le skill triggerano da sole. Se vuoi forzarne una: /skill-name.

🔵 Custom Multiverbe (9) · created sessioni 9 Mag · count corretto audit 2026-05-17

Skill	Trigger keywords	Quando
`multiverbe-design-system`	design, layout, styling, liquid glass, Osservatorio, Nebula Glass, restyle	qualsiasi componente UI nuovo o modifica · skill rinominata 2026-05-14 (era `multiverbe-design`)
`multiverbe-r3f-patterns`	NucleoSphere, R3F, three.js, pianeti, WebGPU	qualsiasi 3D dashboard
`multiverbe-sector-onboarding`	onboarding, settore, 5 macro, AI fallback, TTFV	onboarding flow + sector preset
`multiverbe-cold-outreach-ch`	cold email, demo, prospect, obiezione, W1-W4, Loris	sales outreach Loris
`multiverbe-ch-compliance`	nDSG, revFADP, GDPR, DPA, privacy policy, cookie banner	qualsiasi data protection topic
`multiverbe-security-ch`	security, OWASP, RLS audit, secret rotation, MFA, pentest	security review code path
`multiverbe-stripe-checkout-edges`	Stripe, Checkout, prorate, dunning, trial, refund, webhook	qualsiasi Stripe edge case
`multiverbe-supabase-rls-patterns`	RLS, workspace_id, multi-tenant, auth.uid(), JWT claims	RLS policy review
`multiverbe-incident-response`	incident, breach, ransomware, RLS leak, post-mortem	response a incidente sicurezza

📑 Slash command custom · invocabili in chat

Comando	Cosa fa
`/launch-checklist`	Verifica 6 dimensioni readiness pre-deadline 10 Giu
`/loris-cadence`	Daily 8.5h cadence sprint W2-W4
`/dr-test`	Quarterly DR test 1h (Supabase PITR + integrity check)
`/incident-start`	Kick off incident response 5-step playbook 4h
`/security-weekly`	Lunedì 9:00 · 35min weekly security check
`/security-monthly`	1° del mese · 85min audit

🟢 Skills cloned GitHub OSS (5)

tailwindcss-skills · Tailwind v4.2 reference
a11y-skill · accessibility audit
r3f-fundamentals · React Three Fiber basics
r3f-postprocessing · post-processing effects
r3f-animation · animations spring/keyframes

🟣 Plugins ufficiali Claude Code (14)

Plugin ufficiali installati: claude-md-management, code-review, coderabbit, feature-dev, frontend-design, init, review, security-review, skill-creator, stripe (best-practices, explain-error, test-cards), vercel (deploy, logs, setup), anthropic-skills (consolidate-memory, docx, humanizer, pdf, pptx, schedule, setup-cowork, skill-creator, xlsx).

Sono auto-disponibili in tutte le sessioni Claude Code.

🟡 Skills baseline (32) · n8n + ads + seo

Per task specialistici (workflow automation, paid ads audit, SEO):

n8n (7): code-javascript, code-python, expression-syntax, mcp-tools-expert, node-configuration, validation-expert, workflow-patterns
Ads (13): ads, ads-audit, ads-budget, ads-competitor, ads-creative, ads-google, ads-landing, ads-linkedin, ads-meta, ads-microsoft, ads-plan, ads-tiktok, ads-youtube
SEO (12): seo-audit, seo-competitor-pages, seo-content, seo-geo, seo-hreflang, seo-images, seo-page, seo-plan, seo-programmatic, seo-schema, seo-sitemap, seo-technical
Misc (3): claude-api, owasp-security, multiverbe-* (sopra)

Skill chains · combinazioni utili

Quando un task tocca più domini, Claude può invocare più skill in sequenza.

Chain	Skill coinvolte	Use case
Onboarding nuovo settore	`multiverbe-sector-onboarding` + `multiverbe-design-system` + `multiverbe-r3f-patterns`	Aggiungere preset settore + UI + 3D
Compliance + privacy	`multiverbe-ch-compliance` + `multiverbe-security-ch`	Privacy policy + RLS audit
GTM Loris	`multiverbe-cold-outreach-ch` + `loris-cadence` + `multiverbe-sector-onboarding`	Cold email settore-specifico
Pre-launch	`launch-checklist` + `multiverbe-security-ch` + `multiverbe-ch-compliance` + `multiverbe-stripe-checkout-edges`	Hard gate 10 Giu readiness
Incident response	`incident-start` + `multiverbe-incident-response` + `multiverbe-supabase-rls-patterns` + `multiverbe-ch-compliance`	Breach 4h target
Quarterly DR	`dr-test` + `multiverbe-security-ch` + `multiverbe-supabase-rls-patterns`	DR test trimestrale

🛡️ Security Toolkit · top install pre-launch

GitHub OSS verified stars per security pre-launch:

OWASP ZAP · DAST scanner
Dependabot · dependency updates auto
Snyk · vuln scanning npm/pip
Sentry · error + perf tracking (già stack)
Supabase get_advisors · RLS performance + lint (MCP tool)

🛠️ Da aggiungere/installare

[ ] Stripe Tax integration (post Series-A)
[ ] Yubikey hardware MFA (founder)
[ ] 1Password Teams (founder + Loris)
[ ] BitLocker device encryption (Gregorio Windows + Loris)
[ ] VPN aziendale (Mullvad o ProtonVPN, post 5 paganti)

Tracking + roadmap: skill multiverbe-security-ch.

Come introdurre una nuova skill

Decidi se serve davvero (skill costano contesto · soglia: useremmo 3+ volte/mese?)
Crea via slash /skill-creator o cloning da GitHub
Test trigger keyword in 2-3 prompts diversi
Aggiungi riga qui + log in execution/decisions-log.md

Ultimo update: 2026-05-17 · audit-batch · (1) Count "(8)" corretto a "(9)" (header non allineato a tabella). (2) Skill multiverbe-design → multiverbe-design-system (rinominata 2026-05-14 per allineamento Osservatorio design system). Precedente 2026-05-09: estratto da BOS §16. Owner: Gregorio Ciurleo (decisione install) · Loris Beraud (segnala skill mancanti per sales)