bbssppllvv
- 12 skills
- 0 followers
- 6 hours ago last updated
- ▌ Refero Design · bbssppllvv bundleDeprecated legacy Refero Design skill package. The canonical skill moved to https://github.com/referodesign/refero_skill. Install it with `npx skills add https://github.com/referodesign/refero_skill --skill refero-design`.
- ▌ Soniox · bbssppllvv bundleIntegrate Soniox speech-to-text and text-to-speech APIs for real-time voice applications, voice agents, and transcription. Use when working with Soniox WebSocket streaming, Soniox REST transcription, Soniox TTS, Soniox SDKs (Python, Node, Web), building voice bots/agents with Soniox, or implementing voice/speech features using Soniox. Triggers on mentions of Soniox API, soniox package, @soniox/node, @soniox/speech-to-text-web, real-time transcription via Soniox, speech-to-text integration, text-to-speech with Soniox, Soniox voice bot, Soniox voice agent, or audio transcription with Soniox.
- ▌ Sayless · bbssppllvv bundleWrite and review product copy, UX microcopy, and marketing text. Use this skill when the user asks to write button text, error messages, empty state copy, landing page headlines, onboarding flows, notification copy, or any product-facing text. Also use when reviewing or improving existing copy for clarity, brevity, and impact. Triggers on: "write copy", "button text", "error message copy", "landing page copy", "review this copy", "UX writing", "microcopy", "marketing headline", "empty state text", "improve this copy", "make this copy better", "copywriting help", "rewrite this", "too wordy", "shorten this text", "notification copy", "onboarding copy", "CTA text", "tooltip text".
- ▌ Crucible · bbssppllvv bundleRun an independent multi-perspective review by spawning a small panel of subagents, each with a sharp role, then synthesizing their findings into evidence-based decisions. Invoke autonomously only when independent judgment is likely to materially change a decision before action: meaningful risk, real uncertainty, competing approaches, a stuck diagnosis, a plan that should be stress-tested, or an explicit request for fresh eyes, multiple reviewers, a devil's advocate, red team, jury, outside opinion, or second opinion. Do not invoke for routine low-risk edits, direct factual answers or commands, mechanical docs/tests, tasks without a reviewable artifact, or when already acting as a Crucible reviewer/subagent.
- ▌ Vercel AI · bbssppllvv bundleComplete Vercel AI ecosystem guide: AI SDK v6 (generateText, streamText, useChat, tools, agents, MCP), AI Gateway (routing, fallbacks, caching), AI Elements (48+ React UI components), Chat SDK (Slack/Discord/Teams bots), Workflow DevKit, Data Stream Protocol, and Security (Vercel Firewall/WAF, BotID bot detection, rate limiting, DDoS, Attack Challenge Mode, prompt injection prevention, cost controls). Use when building AI chat UIs, agents, streaming, tool calling, or working with ai/@ai-sdk packages. Also use when securing AI apps: Firewall rules, BotID, @vercel/firewall rate limiting, bot blocking, geo-blocking, IP blocking, AI endpoint protection, Spend Management, or any Vercel security for AI products.
- ▌ Fluidaudio · bbssppllvv bundleIntegrate FluidAudio Swift SDK for fully local, on-device audio AI on Apple platforms. FluidAudio provides ASR (NVIDIA Parakeet TDT models), speaker diarization, VAD, and TTS — all running on Apple Neural Engine via CoreML with zero cloud dependency. Use when implementing: (1) offline/local speech-to-text transcription on macOS/iOS, (2) real-time streaming ASR with end-of-utterance detection, (3) speaker diarization (who spoke when), (4) voice activity detection, (5) text-to-speech synthesis, (6) any FluidAudio SDK integration, FluidInference models, or Parakeet CoreML models. Triggers on mentions of FluidAudio, FluidInference, parakeet-tdt, on-device transcription with CoreML, local ASR on Apple Silicon, or offline speech recognition for macOS/iOS apps.
- ▌ Openrouter · bbssppllvv bundleComplete OpenRouter API documentation reference. Use when working with OpenRouter — an AI model routing platform that provides a unified API for accessing 300+ LLMs (OpenAI, Anthropic, Google, Meta, Mistral, etc.). Triggers: OpenRouter API integration, model routing, provider selection, multi-model fallbacks, OpenRouter SDK (TypeScript/Python), OpenRouter chat completions, OpenRouter embeddings, OpenRouter structured outputs, OpenRouter tool calling, OpenRouter streaming, BYOK (Bring Your Own Key), OpenRouter OAuth, OpenRouter guardrails, prompt caching via OpenRouter, OpenRouter billing/credits, model variants (:free, :extended, :nitro, :thinking, :online), Broadcast observability integrations, Zero Data Retention, or any task involving openrouter.ai API endpoints.
- ▌ Figma Design · bbssppllvv bundleComprehensive Figma guide for AI agents working with design files, design-to-code, canvas automation, Plugin API scripts, MCP tools, variables, components, effects, CSS Grid, REST API, Code Connect, and design-system extraction. Use when reading, creating, editing, or validating Figma designs; building Figma plugins or scripts; setting up official/community Figma MCP servers; comparing Figma capabilities with Paper or Pencil; or handling Figma Design Agent, figma.com, Figma Desktop, Figma REST API, webhooks, Slides, Buzz, codegen, or design token workflows.
- ▌ Product Design · bbssppllvv bundleProduct design best practices — typography, color, spacing, motion, icons, accessibility, and craft details for professional web/mobile UI. Prevents generic "AI slop" designs through intentional, research-grounded decisions. Use when: (1) designing or building UI screens, dashboards, landing pages, or components, (2) reviewing UI for quality and polish, (3) setting up a design system (tokens, scales, palettes), (4) implementing CSS/frontend with professional-grade craft, (5) any task involving visual design quality, typography choices, color palettes, animation/motion, icon systems, or accessibility. Triggers on: UI design, product design, design system, typography, color palette, spacing system, motion/animation, icons, dark mode, accessibility, anti-AI-slop, design review, visual polish, CSS tokens, responsive design.
- ▌ Agent Engineering · bbssppllvv bundleBuild production-grade AI agents and agentic workflows: agent loops, tool calling, Agent-Computer Interface (ACI) design, harness/scaffolding, context engineering, error recovery, permissions, security threat modeling, multi-agent orchestration, cost controls, eval harnesses, production runbooks, and API/version watchlists. Use when designing or implementing LLM agents, coding assistants, Claude Code/Codex/Cursor/Devin-style systems, ReAct loops, tool-use workflows, Claude Agent SDK, Vercel AI SDK ToolLoopAgent, OpenAI Agents SDK, or any system where an LLM reasons and acts through tools.
- ▌ Agent Folder Init · bbssppllvv bundleAdd or repair .agents/ project context for an existing repo. Use for AI agent documentation, session tracking, task management, and coding standards; do not use as the primary new-product scaffold.
- ▌ Polar Integration · bbssppllvv bundleIntegrate Polar payments, subscriptions, and checkout into web projects. Use when asked to add payments via Polar, set up Polar checkout, configure Polar webhooks, create Polar products, integrate Polar SDK, set up customer portal, add subscription billing with Polar, or any task involving polar.sh payment platform. Triggers on mentions of Polar payments, Polar checkout, Polar webhooks, Polar subscriptions, @polar-sh/sdk, @polar-sh/nextjs, @polar-sh/checkout.