Menu
Replace me. One sentence on what this skill does and when to use it.
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Based on SOC occupation classification
Model parametric 3D geometry as code and edit mesh files programmatically. OpenSCAD (.scad) is the core for scripting solid models with named parameters; pair it with a Python mesh toolkit (trimesh/manifold3d) to inspect, repair, measure, scale, boolean, convert, and batch-generate STL/3MF — then optionally slice and print (Bambu Lab / OrcaSlicer). Use whenever the user wants to create, generate, design, model, script, parametrize, modify, repair, convert, or batch-produce a 3D model, part, or mesh: write or improve OpenSCAD/.scad code, turn a description into an STL or 3MF, make a part with adjustable/customizable dimensions, fix a non-watertight or non-manifold STL, scale/cut/boolean/convert an existing STL, render many variants, or design something to 3D print. Trigger even if they only say 'OpenSCAD', '.scad', 'STL', '3MF', 'edit/repair this STL', 'parametric model', 'CAD', 'make me a bracket/box/holder/knob', 'design for 3D printing', 'Bambu', 'Fusion 360', or 'Blender'. Bridges Fusion 360 and Blender.
Validate any Agent Skill against the agentskills.io specification plus project- or skill-local rule overrides. Use when auditing a skill folder for compliance, preparing a skill for distribution, checking PR-changed skills in CI, enforcing custom frontmatter rules, or verifying that a skill's `metadata`, `compatibility`, body length, and file references all meet the spec. Every rule is config-driven — point the validator at a YAML config to customize allowed/required/forbidden frontmatter fields, regex patterns for metadata values, timestamp formats, body thresholds, and placeholder bans. Triggers on: 'validate skill', 'check skill compliance', 'audit SKILL.md', 'lint skill folder', 'verify skill against agentskills.io', 'is this skill spec-compliant'.
Implement Better Auth (better-auth.com) — the framework-agnostic TypeScript authentication library — across ANY stack: Next.js, SvelteKit, Nuxt, SolidStart, Remix/React Router, Astro, TanStack Start, Express, Hono, Elysia, Fastify, or raw Node. Covers the betterAuth() server instance, database adapters (Drizzle, Prisma, Kysely/raw pg·mysql·sqlite, MongoDB), the CLI (generate/migrate), the client SDK (React, Vue, Svelte, Solid, vanilla), email/password + social/OAuth login (Google, GitHub, Apple…), account linking, sessions, and plugins (2FA, magic link, email OTP, passkey, username, organization multi-tenancy, admin, API keys, JWT/bearer, RBAC via access control). Use whenever the user adds, configures, scaffolds, or debugs Better Auth, sign-in/sign-up, sessions, OAuth, 2FA, multi-tenant orgs, or role-based permissions in a TS/JS app — even if they only say 'better auth', 'add login', or 'auth'. Ships a preflight, a scaffolder, and a setup doctor.
Build, scaffold, and integrate Quo — formerly OpenPhone — telephony into a full-stack app: send & receive SMS/text messages, list and analyze calls (recordings, AI summaries, transcripts, voicemails), manage contacts, conversations, tasks, users, and phone numbers via the Quo REST API (https://api.quo.com/v1), plus signature-verified webhooks. Ships a one-command scaffolder that emits a runnable integration (Express + API client + send-SMS route + a signature-verifying webhook receiver), a preflight checker, a tested webhook verifier (Node + Python), and an importable client. Use whenever the user wants to add, build, or scaffold SMS/text messaging, programmatic calls, an OpenPhone or Quo integration, two-way SMS, call transcripts/summaries, contact sync, or Quo webhooks — even if they only say 'OpenPhone', 'Quo', or 'send a text from my app'. Bakes in the raw-API-key auth gotcha (no Bearer token), A2P 10DLC registration, the 202/one-recipient/E.164 send rules, and the whsec_ webhook signing scheme.
Build, scaffold, and integrate MoonPay crypto on-ramp (buy) and off-ramp (sell) payments into a website or app — the embeddable Widget (@moonpay/moonpay-js / -react / -node) and the headless Platform + Frames product (@moonpay/platform-sdk-web). Ships a generator (scripts/scaffold-moonpay.mjs) that emits a complete, runnable integration — server-side URL signing, the frontend widget, a verified webhook handler, env, and README — so you stand up a working on-ramp/off-ramp in one command, then customize. Use whenever the user wants to add, build, generate, or scaffold fiat↔crypto buy/sell, a crypto checkout or 'shopping-style' payment processor, embed a buy/sell widget, integrate MoonPay, do crypto onboarding/KYC, sign a MoonPay widget URL, verify MoonPay webhooks, or build headless/co-branded Apple Pay / Google Pay / card crypto payments — even if they only say 'MoonPay', 'on-ramp', 'off-ramp', or 'buy crypto in my app'. Also bundles tested signing + webhook scripts and source-cited references.
Generate and edit production-ready, watermark-free image assets with Google's Nano Banana models — gemini-3.1-flash-image (Nano Banana 2) and gemini-3-pro-image (Nano Banana Pro) — via the paid Gemini API and the gemini-cli nanobanana extension. Use when the user wants to create, generate, make, draw, design, edit, restore, or iterate on any image, picture, photo, graphic, illustration, logo, or visual asset; mentions Nano Banana, Gemini image generation, text-to-image, or AI images; complains about the Gemini sparkle watermark or wants a watermark removed from an image; or asks for app icons, favicons, hero banners, OG/social cards, YouTube thumbnails, marketing creatives, diagrams, seamless patterns, sprite sheets, or consistent characters. Bundles preflight, a direct-API generator with aspect-ratio and 512/1K/2K/4K control, gemini-cli setup, prompting guidance, and per-asset production recipes.
| name | example-skill |
| description | Replace me. One sentence on what this skill does and when to use it. |
Replace this body with the real instructions for the capability. It is delivered
to the agent ONLY when it calls load_skill("example-skill") (or reads
skill://example-skill), so it costs zero context until then. Write it like you'd
write any SKILL.md: imperative steps, an explicit output contract, and the why.
Tip: add as many registry/skills/<id>/SKILL.md folders as you need — they appear
in list_skills automatically, with no / menu entry and no always-on cost.
