This skill helps users automatically extract YouTube video transcripts and metadata via the BrowserAct API. The Agent should proactively apply this skill when users express needs like extracting full transcript from a specific YouTube video, getting subtitles and metadata for video content analysis, gathering video titles and likes counts, summarizing YouTube videos without watching them, collecting channel details from a video URL, tracking transcript automation for specific videos, scraping YouTube subtitles for internal knowledge bases, fetching full video content for AI summarization pipelines, downloading structured transcripts from YouTube links, analyzing video text content for media research, monitoring video publisher information and channel links, or building datasets from YouTube video transcripts.

Rich artifact panel for AI chat — code, HTML, Mermaid diagrams, tables, and markdown render in a resizable side panel. Use this skill when the user says "add artifacts", "add code preview", "add side panel", "artifact panel", or "rich output".

Generate AI videos on RunComfy via the `runcomfy` CLI — a smart router across the full video-model catalog: HappyHorse 1.0 (Arena #1, native in-pass audio), Wan-AI Wan 2-7 (open weights, audio-driven lip-sync), ByteDance Seedance v2 / 1-5 / 1-0 (multi-modal cinematic), Kling 3.0 / 2-6, Google Veo 3-1, MiniMax Hailuo 2-3, ByteDance Dreamina 3-0. Covers text-to-video (t2v), image-to-video (i2v), and Veo's video-extend endpoint. The skill picks the right model for the user's intent (Arena-#1 quality, multi-shot character identity, in-pass audio, cinematic motion, fastest path, sub-15s clip, longest duration) and ships each model's documented prompting patterns plus the minimal `runcomfy run` invoke. Triggers on "generate video", "make a video", "text to video", "t2v", "image to video", "i2v", "animate", "AI video", "make X move", "video from prompt", "video from image", or any explicit ask to produce a video clip from prompt or still.

Use when low-latency realtime speech recognition is needed with Alibaba Cloud Model Studio Qwen ASR Realtime models, including streaming microphone input, live captions, or duplex voice agents.

Use when cloning voices with Alibaba Cloud Model Studio Qwen TTS VC models. Use when creating cloned voices from sample audio and synthesizing text with cloned timbre.

Audio-only room with LiveKit — circular avatar grid, speaking indicators, raise hand queue. Twitter Spaces / Clubhouse style. Use this skill when the user says "add audio room", "audio chat", "voice room", "spaces", "clubhouse", or "audio-only room".

Build forms with validation in B2C Commerce using SFRA patterns. Use this skill whenever the user needs to create a storefront form (checkout, registration, profile edit, address, contact), define form fields in XML, handle form submission in a controller, add field validation rules, or render forms in ISML templates. Also use when they mention form XML definitions, server.forms.getForm, form groups/actions/lists, or CSRF protection on form posts — even if they just say "I need a registration form" or "add validation to checkout".

Generate presentations, documents, webpages, and social content using Gamma AI. Supports card dimensions, headers/footers, sharing, multiple image sources, and PDF/PPTX export.

Build, configure, and troubleshoot GraalVM Native Image applications. Use this skill for native-image CLI builds, Maven or Gradle Native Build Tools, reachability metadata, and build or run time issue diagnosis.

Build, configure, and troubleshoot GraalVM Native Image applications. Use this skill for native-image CLI builds, Maven or Gradle Native Build Tools, reachability metadata, and build or run time issue diagnosis.

Swift Charts patterns — marks, axes, selection, styling, composition, Chart3D, accessibility, and Audio Graph. Use when building or improving charts in SwiftUI.

Use this skill when loading assets in Phaser 4. Covers the Loader plugin, loading images, spritesheets, atlases, audio, JSON, tilemaps, bitmap fonts, and tracking load progress. Triggers on: preload, this.load, asset loading, spritesheet, atlas, load progress.

Create mathematical animations and visualizations using Manim (ManimCE - Community Edition). Use this skill when users want to build Manim visualizations, create math animations, animate equations, graphs, geometric proofs, 3D objects, or any programmatic video animation. Triggers on requests mentioning "manim", "mathematical animation", "animate equation", "visualize algorithm", "create animation of", "3D visualization", or building explanatory math videos.

Generate, edit, and compose images using Gemini Nano Banana models via portable Python scripts. Handles authentication via API Key or Vertex AI environment variables. Available parameters: prompt, model, aspect-ratio, safety-filter-level. Always confirm parameters with the user or explicitly state defaults before running.

Reference for nimble extract command. Load when fetching URLs or scraping pages. Contains: render tiers 1-3, all flags, browser actions, network capture, parser schemas, geo targeting, async, parallelization.

Generate and edit images using NVIDIA FLUX models. Use when user asks to generate images, create pictures, edit photos, or modify existing images with AI. Supports text-to-image generation and image editing with text prompts.

Create and render OpenSCAD 3D models. Generate preview images from multiple angles, extract customizable parameters, validate syntax, and export STL files for 3D printing platforms like MakerWorld.

MUX video infrastructure — live streaming, VOD uploads, asset management, and webhook processing. Use this skill when the user says "add mux", "setup video streaming", "add live streaming", "setup mux", "video uploads", or "stream-mux".

Dose-response / concentration-response curve fitting — IC50, EC50, Hill slope, Emax/Emin efficacy, and relative potency from paired concentration vs response data (enzyme/cell assays, drug screening, agonist/antagonist pharmacology). Fits the 4-parameter logistic (Hill sigmoidal) model. Use when you have concentrations + responses and need a potency value, to compare two compounds' potency, or to judge curve quality. NOT for image-derived dose-response (use tooluniverse-image-analysis) and NOT for survival/regression (use tooluniverse-statistical-modeling).

Loading and using pretrained models with Hugging Face Transformers. Use when working with pretrained models from the Hub, running inference with Pipeline API, fine-tuning models with Trainer, or handling text, vision, audio, and multimodal tasks.

Deploy applications and websites to Vercel. Use when the user requests deployment actions like "deploy my app", "deploy and give me the link", "push this live", or "create a preview deployment". Do NOT use for deploying to Netlify, Cloudflare, or Render (use their respective skills).

Generate, edit, or understand images using Google Gemini models. Use when asked to "generate an image", "create a picture", "edit an image", "modify this photo", "describe an image", "what's in this image", "analyze this photo", "image to text", "text to image", "explain this screenshot", "visual question answering", or any task involving image generation, image editing, or visual understanding.

Transforms product documentation into sales-ready narrative packages combining story explanation, charcoal gestural sketch art, and talking points. Pipeline: extract narrative arc → determine emotional register (wonder, determination, hope) → derive visual scene → generate assets. Three workflows: CreateSalesPackage (full pipeline — narrative + charcoal sketch visual + key talking points), CreateNarrative (story only, 8-24 numbered points in first-person conversational voice, captures why-it-matters not what-it-does), CreateVisual (charcoal gestural sketch with transparent background for presentation versatility). Charcoal gestural sketch is the mandatory visual style — minimalist composition with breathing space. Output is tied directly to what's being sold — clear, succinct, effective. Integrates StoryExplanation (for narrative arc extraction) and Art essay-art workflow (for visual generation) internally. USE WHEN: sales, proposal, pitch deck, value proposition, sales narrative, sales deck, sales package, turn this into a pitch, create a sales story, sales materials, product pitch, transform docs to sales, sales script. NOT FOR Hormozi $100M frameworks, value equation, irresistible offer, or VOC mining (use _SALESHORMOZI). NOT FOR standalone diagrams or illustrations (use Art).

Use when tasks require all-in-one multimodal understanding or generation with Alibaba Cloud Model Studio Qwen Omni models, including image-plus-audio interaction, voice assistants, and realtime multimodal agents.

Write blog posts in Dhruv Baldawa's distinctive voice - conversational yet analytical, grounded in personal experience, with clear structure and practical insights optimized for Substack. Use when writing or revising draft.md, translating ideas from braindump into polished prose.

Package Claude Code as an OCI image using mise and deploy it as a kubernetes-sigs/agent-sandbox SandboxTemplate workload. Use when building a mise-driven Dockerfile for Claude Code, configuring the ~/.claude named volume for session persistence, injecting ANTHROPIC_API_KEY per SandboxClaim, applying per-pod egress allowlists for api.anthropic.com, or running Claude Code headlessly inside a Kata or gVisor isolated pod.

Build interactive debugging interfaces that reveal internal system behavior. Use when asked to "help me understand how this works", "show me what's happening", "visualize the state", "build a debug view", "I can't see what's going on", or any request to make opaque system behavior visible. Applies to state machines, data flow, event systems, algorithms, render cycles, animations, CSS calculations, or any mechanism with hidden internals.

This skill should be used when the user wants a directed graph — a DAG, workflow, pipeline, or dependency tree — drawn as a plain-text terminal picture: "draw this graph", "sketch the DAG", "diagram these dependencies", "render as ASCII", or turning a Workflow script's pipeline()/parallel() structure into a diagram (even when "ASCII" is never said). Renders layered box-art, zero-dependency, upgrading to graph-easy when installed.

This skill should be used when users have audio interview recordings to transcribe, need to convert PDF documents, mentions 'import data', 'transcribe', 'convert', or is starting data preparation for Stage 1 or Stage 2.

Browser-based video processing with mediabunny — read, write, convert MP4/WebM/MKV using WebCodecs API. Server integration for storage and metadata tracking.

Edit and enhance images and videos with AI via muapi.ai — prompt-based editing, upscaling, background removal, face swap, lipsync, video effects, and more

Generate all required PWA assets from a single 1024x1024px image. This skill creates app icons, favicons, Apple touch icons, maskable icons, badge icons, and placeholder screenshots for PWAs. Use when developers need to prepare PWA manifest assets, generate multiple icon sizes from a source image, or create a complete set of PWA-compliant images.

Remove marcas d'água de imagens, especialmente de geradores de IA como Google Gemini, Nano Banana, DALL-E, Midjourney etc. Use esta skill quando o usuário pedir para remover watermark, marca d'água, logo de IA, selo de gerador, ou limpar imagem gerada por IA. Também ativa quando o usuário menciona 'marca d'água', 'watermark', 'logo do gemini', 'selo da IA' ou quer limpar cantos de imagens de IA. Inclui setup automático em virtualenv isolado com OpenCV.

Transform raw podcast transcripts into polished, readable documents. This skill removes filler words, fixes grammar, adds structure, and preserves authentic voice. Use when processing podcast transcripts, interview recordings, video captions, or any spoken content that needs cleanup while maintaining 100% fidelity to meaning.

Generate professional walkthrough videos from app screenshots or live sites using Remotion. Smooth transitions, zoom effects, text overlays, and optional voiceover narration. Produces MP4 videos for demos, product showcases, or documentation. Triggers: 'walkthrough video', 'demo video', 'product video', 'create a video walkthrough', 'remotion video', 'screen recording', 'app demo', 'showcase video', 'generate video from screenshots'.

Download YouTube videos with quality presets. Use for: download youtube, yt download, video download, youtube to whatsapp, youtube mp3.

Use when visual reasoning is needed with Alibaba Cloud Model Studio QVQ models, including step-by-step image reasoning, chart analysis, and visually grounded problem solving.

Smoke test for aliyun-solution-article-illustrator. Validate solution-skill structure, compatibility-oriented documentation, prompt-before-image workflow, backend recommendation policy, and final Markdown output artifacts.

Use when Alibaba Cloud Model Studio Wan video editing models are needed for style transfer, keyframe-controlled editing, or animation remix workflows.

Authors and debugs AWS CodePipeline + CodeBuild workflows — pipeline v1 vs v2 (triggers, variables), source providers via CodeStar Connections, artifact handoff, buildspec.yml authoring, IAM service roles, ECR pull permissions, VPC build environments, S3/local caching strategies, Lambda invoke action callback pattern, and manual approval setup. Use when working with AWS CodePipeline, AWS CodeBuild, buildspec.yml, CodeStar Connections, pipeline service roles, build VPC config, or "CodeBuild can't pull image" / "Lambda action hangs" debugging.

Compresses images to WebP (default) or PNG with automatic tool selection. Use when user asks to "compress image", "optimize image", "convert to webp", or reduce image file size.

Compresses images to WebP (default) or PNG with automatic tool selection. Use when user asks to "compress image", "optimize image", "convert to webp", or reduce image file size.

Writes technical blog posts for developers: researches current state, drafts outlines for approval, then writes in an authentic voice with a clear "aha moment", strong hook, and concrete code examples. Make sure to use this skill whenever the user says "write a blog post", "draft a post about", "blog post on", "technical article", or runs /blog. Also triggers when the user wants to publish something they've learned or share an opinion — even if they just say "I want to write about X".

Generate or edit images using AI models (FLUX, Gemini). Use for general-purpose image generation including photos, illustrations, artwork, visual assets, concept art, and any image that isn't a technical diagram or schematic. For flowcharts, circuits, pathways, and technical diagrams, use the scientific-schematics skill instead.

Remove signs of AI-generated writing from text. Use when editing or reviewing text to make it sound more natural and human-written. Based on Wikipedia's comprehensive "Signs of AI writing" guide. Detects and fixes patterns including: inflated symbolism, promotional language, superficial -ing analyses, vague attributions, em dash overuse, rule of three, AI vocabulary words, passive voice, negative parallelisms, and filler phrases.

Inworld TTS API. Covers voice cloning, audio markups, timestamps. Use when integrating Inworld text-to-speech, cloning voices, adding audio markups (SSML-like), or aligning viseme timestamps. Keywords: Inworld, text-to-speech, TTS, voice cloning, visemes.

---

Queue job management patterns, processors, and async workflows for video/image processing

Create platform-optimized social media posts for LinkedIn, Instagram, or X/Twitter with author voice adaptation

Generate SRT subtitles from video/audio with translation support. Transcribes Hebrew (ivrit.ai) and English (whisper), translates between languages, burns subtitles into video. Use for creating captions, transcripts, or hardcoded subtitles for WhatsApp/social media.

Use this skill when reading video-analytics metrics, incidents, alerts, and sensor data via the VA-MCP server (port 9901). Not for live VLM or incident-range narrative reports.

Extract YouTube video transcripts with metadata and save as Markdown to Obsidian vault. Use this skill when the user requests downloading YouTube transcripts, converting YouTube videos to text, or extracting video subtitles. Does not download video/audio files, only metadata and subtitles.

Comprehensive academic textbook chapter writing system for medical/scientific content. Use when the user wants to: (1) Write a full textbook chapter (5,000-15,000 words) on any medical/scientific topic, (2) Generate a detailed table of contents with section word counts, (3) Research topics via PubMed MCP and compile 20-30 references, (4) Write section-by-section with proper citations in Vancouver format, (5) Create publishable academic content with Eric Topol-inspired voice and authentic human prose, (6) Get approval at TOC stage before writing begins, (7) Export well-structured chapters for textbook publication.

Codex Pet generator on RunComfy. Build a Codex-compatible Codex Pet spritesheet.webp + pet.json from a single reference image, drop it into `${CODEX_HOME:-$HOME/.codex}/pets/<name>/` and Codex picks it up as a custom Codex Pet next to the 8 built-ins. This skill produces the exact Codex Pet atlas Codex expects (1536x1872 PNG/WebP, 8 cols x 9 rows, 192x208 cells, 9 animation states — idle, running-right, running-left, waving, jumping, failed, waiting, running, review). Calls OpenAI GPT Image 2 edit ONCE via the local RunComfy CLI as `runcomfy run openai/gpt-image-2/edit` to produce a canonical Codex Pet pose, then assembles all 9 animation rows programmatically with ImageMagick micro-transforms — no Codex Pro, no `$imagegen`, no OPENAI_API_KEY required, only RUNCOMFY_TOKEN. Triggers on "codex pet", "create codex pet", "make codex pet", "hatch codex pet", "/hatch image", "desktop pet codex", "codex pets", "spritesheet.webp", or any explicit ask to build a custom pet for OpenAI Codex.

Codex Pet generator on RunComfy. Build a Codex-compatible Codex Pet spritesheet.webp + pet.json from a single reference image, drop it into `${CODEX_HOME:-$HOME/.codex}/pets/<name>/` and Codex picks it up as a custom Codex Pet next to the 8 built-ins. This skill produces the exact Codex Pet atlas Codex expects (1536x1872 PNG/WebP, 8 cols x 9 rows, 192x208 cells, 9 animation states — idle, running-right, running-left, waving, jumping, failed, waiting, running, review). Calls OpenAI GPT Image 2 edit ONCE via the local RunComfy CLI as `runcomfy run openai/gpt-image-2/edit` to produce a canonical Codex Pet pose, then assembles all 9 animation rows programmatically with ImageMagick micro-transforms — no Codex Pro, no `$imagegen`, no OPENAI_API_KEY required, only RUNCOMFY_TOKEN. Triggers on "codex pet", "create codex pet", "make codex pet", "hatch codex pet", "/hatch image", "desktop pet codex", "codex pets", "spritesheet.webp", or any explicit ask to build a custom pet for OpenAI Codex.

Remove LLM-isms and AI writing patterns from text. This skill should be used when editing prose to sound less like AI output — removing overused words, fixing structural tells, and restoring natural human voice. Triggers: "de-slop", "remove AI writing", "humanize this", "sounds too AI", "LLM-isms", "AI slop", or when reviewing text that reads like chatbot output.

Save a live webpage as a high-fidelity PDF that preserves the original layout AND every image (including lazy-loaded ones) using the agent-browser CLI. Use this whenever the user asks to "download this page as PDF", "save this article", "archive this URL", "fetch this page for reference", or otherwise wants a local PDF of a web page that looks like the browser version. Especially important on modern JS-heavy sites (engineering blogs, Next.js sites, anything with IntersectionObserver lazy loading) where naive `chrome --headless --print-to-pdf` or a bare `agent-browser pdf` produces blank rectangles or broken-image placeholders. Trigger this skill even when the user does not name the tool - any request to capture a webpage's full visual content as a PDF on disk should pull this in. For reader-mode/article-only output (no nav, no footer, no manual trimming) prefer percollate instead - see "When NOT to use this".

Model-family-specific prompt craft for fal.ai endpoints. Trigger when the user mentions a specific model family by name and asks how to prompt it ("how do I prompt Kling", "GPT Image 2 prompt structure", "Happy Horse tips"), or when prompts to a routed endpoint keep coming back generic and the family's known nuances should be applied. For endpoint selection ("which model for X"), use `fal-models-catalog` instead. This skill is about how to talk to a model once it has been chosen.

Remove signs of AI-generated writing from text. Use when editing or reviewing text to make it sound more natural and human-written. Based on Wikipedia's comprehensive "Signs of AI writing" guide. Detects and fixes patterns including: inflated symbolism, promotional language, superficial -ing analyses, vague attributions, em dash overuse, rule of three, AI vocabulary words, negative parallelisms, and excessive conjunctive phrases. Credits: Original skill by @blader - https://github.com/blader/humanizer

Remove signs of AI-generated writing from text. Use when editing or reviewing text to make it sound more natural and human-written. Based on Wikipedia's comprehensive "Signs of AI writing" guide. Detects and fixes patterns including: inflated symbolism, promotional language, superficial -ing analyses, vague attributions, em dash overuse, rule of three, AI vocabulary words, negative parallelisms, and excessive conjunctive phrases. Credits: Original skill by @blader - https://github.com/blader/humanizer

Server-side room recording with LiveKit Egress — composite and track-based recording, S3 storage, recording metadata in Postgres. Use this skill when the user says "add recording", "record room", "record video", "record meeting", "egress recording", or "save recording".

This skill should be used when the user asks to "generate video", "create video", "make a video", "animate image", "image to video", "video with audio", "talking video", or needs AI-powered video generation with synchronized audio, lip-syncing, and cinematic camera controls.

Download videos from YouTube and other sites using yt-dlp. Use this skill whenever the user wants to download, save, grab, or rip a video from a URL. Supports quality selection, subtitle download/embedding, format options, and audio extraction. Triggers on any video download request, even if the user just pastes a video link and says "download this".

This skill should be used when the user asks about "controllers", "actions", "params", "strong parameters", "filters", "before_action", "after_action", "callbacks", "respond_to", "render", "redirect_to", "flash messages", "sessions", "cookies", "request handling", "routing constraints", or needs guidance on building Rails controllers.

Add images, diagrams, and video embeds to an existing blog post.

Bulk-retouch a folder of portrait photos using Adobe tools — designed for wedding photographers and event photographers who need fast, walk-away batch processing. Use this skill when the user says "retouch my photos", "batch process these portraits", "process my wedding photos", "clean up this folder of images", "run my headshots through Adobe", or uploads/selects a folder of photos and wants them polished and ready to review. Automatically applies auto-straighten, auto-tone, and auto-light to every image. Outputs a preview grid and download folder. Access: 🔐 Signed-In required | Gen AI: ❌

End-to-end AI video generation - create videos from text prompts using image generation, video synthesis, voice-over, and editing. Supports OpenAI DALL-E, Replicate models, LumaAI, Runway, and FFmpeg editing.

Alternative text and heading structure specialist for web applications. Use when building or reviewing any page with images, icons, SVGs, videos, figures, charts, or heading hierarchies. Covers meaningful vs decorative images, complex image descriptions, heading levels, document outline, and landmark structure. Can analyze images visually, compare existing alt text against image content, and interactively suggest appropriate alternatives. Applies to any web framework or vanilla HTML/CSS/JS.

Angular v19-v21 features: linkedSignal, resource/httpResource, signal forms, route-level SSR render modes, incremental hydration, zoneless change detection, Vitest migration, and Angular Aria components.

Markdown AST operations for Jerry documents. Provides structured parse, query, modify, validate, and render operations on Jerry markdown files via the jerry ast CLI.

BlitzReels API umbrella skill: auth, endpoint discovery, onboarding, public API caveats, and links to specialized BlitzReels skills. Use this whenever a user asks an agent to call BlitzReels, complete BlitzReels onboarding, debug BlitzReels API behavior, verify endpoints, or work with BlitzReels API keys before moving into clipping, video editing, caption themes, or carousels.

Author Mermaid diagrams that render cleanly when pasted into Excalidraw's "Mermaid to Excalidraw" feature. Use when the user asks for a diagram, flow chart, sequence diagram, class diagram, architecture diagram, or specifically mentions Excalidraw, Mermaid, drawing, sketching, "draw this", "make a diagram", or "paste into excalidraw".

Write, explain, and debug Metal GPU code including shaders (vertex, fragment, compute), render pipelines, compute pipelines, buffer/texture management, and Metal 4 APIs. Use this skill whenever the user mentions Metal, GPU programming, shaders, MSL (Metal Shading Language), render passes, compute kernels, MTLDevice, MTLCommandBuffer, MTLRenderPipelineState, or any Apple GPU/graphics programming topic. Also trigger when the user wants to do parallel computation on Apple devices, write GPU-accelerated code, or work with Metal Performance Shaders, MetalFX, MetalKit, or Compositor Services. Covers iOS, macOS, iPadOS, tvOS, and visionOS.

Post-process raw screen recordings by removing silent segments and applying speed adjustments. Uses FFmpeg-based Python scripts to optimize video pacing automatically.

Query Stream data and run CLI operations against Chat, Video, Feeds, and Moderation: list channels, list calls, show flagged messages, find users, query any Stream resource. Run stream api / stream config / stream auth commands. Install the Stream CLI binary. Use when the user gives operational verbs ('list', 'show', 'find', 'check', 'query') with Stream nouns, or invokes the CLI literally.

Search Xiaohongshu (RedNote / xhs) notes by keyword and return a paginated list with title, author, engagement stats (likes, collects, comments), cover image URL, and xsecToken for detail lookup. Use when user mentions find notes on xiaohongshu, search rednote, search xhs, scrape xiaohongshu search, xiaohongshu keyword search, rednote post search, xhs search results, monitor xiaohongshu topics, KOL content discovery via xiaohongshu, xiaohongshu note list, rednote scrape, xhs data collection, collect xiaohongshu posts, xiaohongshu topic search, xiaohongshu content monitoring, rednote post list, xhs keyword scrape.

Implement text-to-speech (TTS) capabilities using the z-ai-web-dev-sdk. Use this skill when the user needs to convert text into natural-sounding speech, create audio content, build voice-enabled applications, or generate spoken audio files. Supports multiple voices, adjustable speed, and various audio formats.

Implement text-to-speech (TTS) capabilities using the z-ai-web-dev-sdk. Use this skill when the user needs to convert text into natural-sounding speech, create audio content, build voice-enabled applications, or generate spoken audio files. Supports multiple voices, adjustable speed, and various audio formats.

Generate images using the internal Google Antigravity API (Gemini 3 Pro Image). High quality, native generation without browser automation.

Transform AI-sounding writing into natural, human prose. Use when creating or editing any written content—articles, essays, reports, emails, social posts, scripts, or marketing copy—to avoid patterns that signal AI authorship. Triggers on requests to "make it sound human," "less AI," "more natural," improve voice/tone, or any writing task where authentic style matters.

Expert knowledge for Azure AI Speech development including troubleshooting, best practices, decision making, limits & quotas, security, configuration, integrations & coding patterns, and deployment. Use when using STT/TTS, custom speech models, Voice Live agents, SSML/avatars, or telephony/LLM integrations, and other Azure AI Speech related development tasks. Not for Azure Communication Services (use azure-communication-services), Azure AI Bot Service (use azure-bot-service), Azure Notification Hubs (use azure-notification-hubs), Azure AI Immersive Reader (use azure-immersive-reader).

Compresses images to WebP (default) or PNG with automatic tool selection. Use when user asks to "compress image", "optimize image", "convert to webp", or reduce image file size.

Compresses images to WebP (default) or PNG with automatic tool selection. Use when user asks to "compress image", "optimize image", "convert to webp", or reduce image file size.

Compresses images to WebP (default) or PNG with automatic tool selection. Use when user asks to "compress image", "optimize image", "convert to webp", or reduce image file size.

Rank content angles by engagement data, competition level, and platform fit. Data-driven angle selection instead of guesswork. Use this skill when the user has a keyword or product and needs to decide WHAT to create, which angle to take, which format to use, or which platform to target. Triggers on: "what angle should I use", "rank content ideas for [keyword]", "best angle for [product]", "which content idea will perform best", "help me pick an angle", "what should I write about", "content angle for [topic]", "rank my content ideas", "which approach will get the most views", "data-driven content planning", "angle ranker", "content scoring", "which hook should I use", "compare these content ideas", "prioritize my content angles", "what video should I make".

Transcribe audio files using ElevenLabs Speech-to-Text (Scribe v2).

Generate and edit images with Gemini API using pure Python stdlib. Zero dependencies - works on locked-down environments where pip/uv aren't available.

Remove signs of AI-generated writing from text. Use when editing or reviewing text to make it sound more natural and human-written. Based on Wikipedia's comprehensive "Signs of AI writing" guide. Detects and fixes patterns including: inflated symbolism, promotional language, superficial -ing analyses, vague attributions, em dash overuse, rule of three, AI vocabulary words, negative parallelisms, and excessive conjunctive phrases.

Generate and edit high-quality images using Gemini 2.5 Flash Image and Gemini 3 Pro Image (Nano Banana). Supports Text-to-Image, Style Transfer, Virtual Try-On, and Character Consistency.

Generate images with Nano Banana Pro via OpenRouter. Use when the user asks for image generation, mentions Nano Banana Pro, Gemini 3 Pro Image, or OpenRouter image generation.

GPU-accelerated data curation for LLM training. Supports text/image/video/audio. Features fuzzy deduplication (16× faster), quality filtering (30+ heuristics), semantic deduplication, PII redaction, NSFW detection. Scales across GPUs with RAPIDS. Use for preparing high-quality training datasets, cleaning web data, or deduplicating large corpora.

GPU-accelerated data curation for LLM training. Supports text/image/video/audio. Features fuzzy deduplication (16× faster), quality filtering (30+ heuristics), semantic deduplication, PII redaction, NSFW detection. Scales across GPUs with RAPIDS. Use for preparing high-quality training datasets, cleaning web data, or deduplicating large corpora.

GPU-accelerated data curation for LLM training. Supports text/image/video/audio. Features fuzzy deduplication (16× faster), quality filtering (30+ heuristics), semantic deduplication, PII redaction, NSFW detection. Scales across GPUs with RAPIDS. Use for preparing high-quality training datasets, cleaning web data, or deduplicating large corpora.

This skill should be used when the user asks to "create a notebook", "add source to notebook", "generate audio overview", "create podcast", "manage NotebookLM", "nlm", "add PDF to notebook", "list notebooks", "summarize sources", "generate study guide", "create FAQ", "briefing document", "chat with notebook", "generate outline", "research a topic", "deep research", or needs to interact with Google NotebookLM via the nlm CLI tool.

Batch-generate images via OpenAI Images API. Random prompt sampler + `index.html` gallery.

Provisions SIP trunks, endpoints, ACLs, credential lists, and phone numbers via the Sinch Elastic SIP Trunking REST API. Use when the user needs SIP connectivity, trunk provisioning, inbound/outbound PSTN voice routing, PBX integration, or SIP-to-PSTN bridging.