michailbul
59 verified skills1 total stars
seedance-higgsfield
Seedance 2.0 video prompt director. Converts plain-text scene descriptions into production-ready bilingual EN+ZH video prompts optimized for the Seedance 2.0 video generator. Handles all Seedance work — action (combat, pursuit, stunts), general (landscapes, journeys, atmosphere), dialogue (confrontations, negotiations, interrogations), and non-narrative commercial work (ad spots, music videos, fashion films, automotive inserts, product shots, pet/character demos, cutaway montages, social reels for TikTok / Reels / YouTube Shorts). Use whenever the user wants to create a Seedance video prompt, mentions Seedance, or describes a cinematic scene for video generation. For NARRATIVE screenplay-integrated work, use seedance-screenwriter instead.
development1
deepgram-transcribe
Transcribe audio files (ogg/mp3/wav/m4a/webm) using the Deepgram Nova-2 API. Use when the user sends a voice message or asks to transcribe audio.
development
brand-guidelines-builder
Builds a complete, reusable brand guidelines document for any company, product, startup, creative project, or idea — from scratch or from a brief. Use this skill whenever a user wants to define, create, or document a brand identity. Triggers include: "create brand guidelines for X", "define our brand", "build a brand identity", "what should our brand look like", "brand colors for my startup", "brand kit", "brand system", "visual identity", "brand document", "style guide", or any request to establish or formalize a brand. Always use this skill when a new project needs a brand foundation before design work begins. Once guidelines are generated, they can be saved as a project-specific skill and reused across all future artifacts.
development
typography-master
Expert web typography consultant. ALWAYS use this skill when working on any web design task that involves choosing fonts, type pairings, typographic hierarchy, or font-related decisions. Triggers include: "pick a font", "what font should I use", "typography for my site", "font pairing", "heading font", "body font", "what typeface fits", or any moment during web page/UI design when fonts come up — even if the user just says "make it look better". When triggered, this skill instructs Claude to recommend a curated shortlist of 4–7 fonts, explain the rationale for each, then build an interactive HTML font picker page with a live hero section so the user can visually compare fonts and make an informed choice. Also covers type scale, line height, letter spacing, and variable font usage for production-ready web typography.
development
image-to-prompt
Reverse-engineer any image into a structured text-to-image prompt using the laniameda cinematic formula (Subject → Scene → Composition → Lighting → Style → Quality → Negative). Use when given an image and asked "give me a prompt", "how do I recreate this", or "reverse engineer this".
testing
seedance-director
Seedance 2.0 scene direction API — converts plain-text scene descriptions into structured bilingual EN+ZH JSON video prompts optimized for Seedance 2.0. Handles action scenes (combat, pursuit, stunts), general scenes (landscapes, journeys, atmosphere), and dialogue scenes (confrontations, negotiations, interrogations) using archetype routing. Outputs pure JSON with no commentary. Use this skill whenever the user wants a production-ready Seedance prompt in JSON format, needs bilingual EN+ZH video prompts, describes an action/fight/chase/dialogue scene for Seedance, wants archetype-routed scene direction, or says "direct this scene", "JSON prompt", "bilingual prompt", "ZH prompt", or "seedance director". This skill is the structured-output counterpart to the other Seedance prompting skills — use it when the user needs machine-parseable JSON output rather than plain-text prompts, or when they need Chinese language prompts alongside English.
development
laniameda-youtube-digest
Watch a YouTube video so Michael doesn't have to. Extract what's actionable — tools, prompts, workflows, "if X then Y" mappings, model-specific techniques — and save to KB. Uses Supadata API for transcripts and vision extraction (on-screen content like ComfyUI node graphs, prompts not spoken aloud). Also handles learning-oriented triggers for AI creatorship tutorials. Keywords: digest video, watch this, youtube extract, video summary, youtube transcript, learn from this, what does this teach, upgrade your skills, ai tutorial, prompting tutorial.
tools
Agentic Website Stack
Build websites with a full agentic layer — not just frontend, but autonomous content publishing, backend dashboard (Kanban CMS), brand corpus as AI brain, block-based generative UI, hero image generation, trend scanning, and Convex cron-driven agents that write + publish on autopilot. Use when the user says "build an agentic website", "website that publishes itself", "autonomous content site", "agentic stack", "website with AI agents", "self-publishing blog", "SEO autopilot site", or wants to add an agentic content layer to any web project.
development
ai-typography
Generate premium branded typography visuals using Nano Banana Pro + font specimen reference images. Replicates the LTX Elements workflow (font consistency via reference image injection) without needing LTX Studio. Use when: user wants to generate a typographic design with a specific font, needs branded text visuals, wants to apply font presets from the Amir Mushich method. Triggers on: "generate typography", "font design", "text visual", "branded text", "type treatment", "font preset", "ai typography", "make this look elite", "typographic visual".
testing
ai-video-prompting
Write AI video prompts at any level — from simple idea prompts to multi-shot cinematic sequences and reference-controlled scenes. Use when Michael asks to generate video prompts, plan a video scene, write dialogue for AI characters, or build a video sequence with Kling, Runway, Veo, or when the target model is not fixed yet. If the target model is explicitly Seedance, prefer `seedance-prompting`.
development
andromeda-messages
Add, update, delete, or list regular note and article nodes in the Andromeda Galaxy app. Also lock/unlock galaxies programmatically. Use when Michael asks to add a message/thought/record to any galaxy (Dreams, Life, Soulmate, Family, Antidreams), add an article/markdown note, list/edit/delete nodes, unlock a galaxy, or check galaxy status. Keywords: add to andromeda, save to galaxy, add to dreams, add to жизнь, unlock your life, разблокируй семью, delete node, update node.
testing
async-page-transitions
GSAP-powered async page transitions with dual-container DOM pattern, clip-path reveals, scale-down crossfades, and SplitText enter animations. Use this skill whenever the user wants smooth page transitions, route-based animations, cinematic page reveals, SPA transitions without a framework, GSAP page animations, clip-path transition effects, or any "page transition" / "route transition" / "page reveal" / "crossfade between pages" request in a web project. Also trigger when user mentions "codrops transitions", "async transitions", or wants to add premium-feeling navigation animations to a site. Works with vanilla JS, React, Next.js, or any frontend stack — adapt the pattern to the framework.
tools
burst-frame-cars
Generate BURST FRAME prompts for automotive reference frame generation using the @kaigani technique. Produces a 20-shot (or 10-shot) Seedance 2.0 Rapid Fire prompt that outputs one video containing consistent, full-resolution car reference frames — then extracts them with ffmpeg. Use this skill whenever the user wants to: generate reference images of a car, create automotive consistency frames for Seedance or Nano Banana, get 20 angles of a car, build a studio shot list for a vehicle, or says "burst frame" for any car or vehicle. Also trigger on: "car reference frames", "20 shots of [car]", "10 shots of [car]", "consistency frames for [car]", "studio prompt for [car]", "automotive burst", "burst frame [car name]", "[car] reference sheet", "generate angles of my car". Works for any car, any environment: studio, road, canyon, mountain, forest, desert, night.
development
car-angle-extractor
Extract shot composition DNA from any car photograph into structured JSON — camera angle, lens, framing, lighting — stripped of car-specific details. Then reuse extracted angles with any car identity to generate new images at scale. Use when: extracting angles from reference photos, building a shot library, batch-analyzing car photography, replicating a great angle with a different car, running extraction pipelines in Freepik or Flora. Triggers: "extract this angle", "steal this composition", "shot DNA", "analyze this car photo", "replicate this shot with my car", "batch extract angles", "car photography analysis", "angle extraction", "build a shot library".
development
carousel-brand-system
Build a complete brand design system tuned for social media carousel production. Interactive workflow: propose color palettes with reasoning, explore font pairings, generate an HTML picker for the user to click what resonates, analyze picks, lock the system in Figma, and produce carousel-format combo cards. Triggers: "build a carousel design system", "create brand colors for carousels", "design system for social media", "carousel brand system", "explore colors for our brand", "pick fonts for carousels", "lock our design system", "brand system for content production", "social media design system", or any request to build a visual identity system specifically for carousel / social media content production. Does NOT trigger for: building individual carousels (use carousel-designer), general brand guidelines docs (use brand-guidelines-builder), or website design systems (use laniameda-brand-design).
tools
character-consistency-character-sheet
Create consistent AI characters using the character sheet method. Use when the user wants character consistency, reference sheets, model turnarounds, or maintaining character identity across multiple generations. Works with Nano Banana Pro/2, Google Flow, Midjourney, and other image models. Triggers on: character sheet, character consistency, reference sheet, model turnaround, identity lock, same character multiple angles, AI character, consistent face.
development
design-inspiration
Loads visual design references to guide any design creation or modification task. ALWAYS use this skill whenever Claude is about to create or modify any design artifact — including UI components, landing pages, web interfaces, dashboards, React/HTML code, Figma outputs, slide decks, or any visual/frontend work. Even if the user doesn't say "use my references" — if a design is being made or changed, consult this skill first. The references folder contains images and notes representing the target aesthetic; Claude must read and internalize them before producing any design output. Also use this skill when the user says "save to gallery", "add to gallery", "save this design", or wants to archive a reference, screenshot, or URL into the laniameda.gallery Design pillar.
development
design-playground
Spin up isolated React playgrounds with Leva control panels to experiment with UI effects, animations, shadows, hover states, and interaction patterns. Use when: 'build me a playground', 'give me knobs for', 'I want to experiment with', 'let me tweak', 'playground for X', 'try this effect', or when user wants to visually dial in a UI technique before integrating into production code. Also triggers when user shares a URL/tweet/CodePen and says 'I want to play with this' or 'let me try this technique'.
development
frame-vfx-stylizer
Convert source video into stylized frame-by-frame mixed-media animation using AI image editing. Three presets: white marker outline, painted background strokes, selective graphic fill. Use when asked to stylize a video frame-by-frame, add stop-motion effects, or create animated sketch overlays.
data-ai
human-copy-standards
Quality gate for all written copy. Kills inflation words, AI-speak, filler, and corporate tone. Enforces human voice: direct, specific, no fluff. Apply before delivering any copy output. Keywords: copy review, writing standards, kill AI speak, human voice, edit copy, no filler.
testing
Laniameda Brand Design
This skill should be used when the user asks to "design a landing page", "create a website design", "build a page in Pencil", "follow the laniameda design system", "use our brand colors", "design for developers", or mentions laniameda brand guidelines, warm editorial style, brutalist-editorial hybrid, or code-savvy developer marketing pages. Provides the complete laniameda design system (colors, typography, shadows, components) and a proven Pencil MCP workflow for building high-quality landing pages.
tools
laniameda-carousel-copy
Write slide-by-slide carousel copy for Laniameda social media posts. Handles the CONTENT side — what goes on each slide, hook writing, copy frameworks, CTA strategy — not the visual design or rendering. Triggers: "write a carousel", "carousel copy", "what should the slides say", "write slides for", "carousel about [topic]", "draft a carousel", "carousel script", "slide copy for", or any request to write the text content for a carousel post. Does NOT trigger for: rendering/exporting carousels (use carousel-designer), building design systems (use carousel-brand-system), or general social posts without carousel format (use social-content).
development
skills/marketing/laniameda-hq-update
--- name: laniameda-hq-update description: Update the Laniameda HQ ground truth docs at ~/work/laniameda/laniameda-hq/ when Michael shares thoughts, ideas, decisions, or direction about the studio — including brand, vision, values, design aesthetic, content standards, agency positioning, projects, or agent protocols. Trigger when: Michael says "add this to the docs", "update the HQ", "lock this in", "write this down for the studio", shares brand or aesthetic opinions, clarifies what we do/don't
testing
laniameda-instagram-carousel-extract
Extract content from Instagram carousel posts (image slides) using browser-use-cloud with visual reading. Read text overlaid on slides, extract prompts, workflows, checklists — and save to KB. Not for reels (video) — use laniameda-instagram-reel-digest for those. Keywords: instagram carousel, extract slides, read carousel, save carousel.
testing
laniameda-instagram-reel-digest
Digest an Instagram reel (video with speech) so Michael doesn't have to watch it. Uses Supadata for transcript first, falls back to browser-use-cloud. Extracts what's said, shown, and actionable. Not for carousel posts (image slides) — use laniameda-instagram-carousel-extract for those. Keywords: instagram reel, digest reel, watch this reel, extract reel, instagram video.
devops
magnetic-dots-background
Implement an interactive magnetic dotted canvas background with ripple effects, flow animation, and mouse attraction physics. Use when building landing pages, hero sections, or any full-screen interactive background in React / Next.js. The surface is achromatic and dark-mode-aware — dots are white on dark, gray on light. Supports click-triggered ripple waves and sinusoidal ambient flow. Drop-in component with full prop control.
tools
mj-nb2-pipeline
Two-stage image pipeline: generate a scene in Midjourney, then enhance realism in Nano Banana (NB2/Pro). Use when the user generates images with Midjourney and then modifies or enhances them with Nano Banana or another image model. Trigger keywords: midjourney + nano banana, MJ to NB2, enhance my MJ output, take my midjourney image into nano banana, realism upgrade, enhance without changing composition.
devops
notion-sync
Sync a task to the laniameda Notion kanban board (upsert by name). Use whenever a task is created/updated/completed in pm/todos/.
tools
parallel-deep-research
ONLY use when user explicitly says 'deep research', 'exhaustive', 'comprehensive report', or 'thorough investigation'. Slower and more expensive than parallel-web-search. For normal research/lookup requests, use parallel-web-search instead.
development
repo-kanban-pm
Install and enforce a lightweight product-management workflow inside a code repo: feature-as-kanban boards, ROADMAP status tracking, branch/PR conventions, and an optional daily OpenClaw cron PM review. Use this skill when (1) starting work in a NEW repository/project that does not yet have a ROADMAP+feature-KANBAN system, (2) you are delegated to spin up a new project and want multi-agent coordination from day one, or (3) you are asked to fix/restructure an existing repo and introduce an organized backlog/feature tracking system. Also use when adding a daily PM audit loop (cron) to keep code + docs + PRs in sync.
development
scroll-framer-motion
Framer Motion / Motion scroll animation skill for React projects. Covers useScroll, useTransform, useSpring, whileInView, parallax, horizontal scroll, sticky sections, text reveals, and staggered animations — all without GSAP. React-idiomatic scroll experiences using Motion + CSS sticky + Lenis. Use when: Framer Motion scroll, Motion scroll, React scroll animation, useScroll, whileInView, React parallax, scroll animation without GSAP. For GSAP-based scroll with ScrollTrigger/pinning/image sequences, use scroll-experience skill instead.
development
seedance-prompting
Write high-performance Seedance 2.0 prompts for NON-NARRATIVE work — ad spots, music videos, fashion films, automotive inserts, product shots, pet/character demos, cutaway montages, social reels (TikTok / Reels / YouTube Shorts), and any cinematic AI video that is not a story scene. Use whenever the target model is explicitly Seedance and the work is standalone or commercial — no dialogue, no scene continuity, no shot-list-into-a-film. Also owns Seedance-specific structured output (JSON, bilingual EN+ZH prompt delivery) for non-narrative briefs. For NARRATIVE work — scenes, short films, screenplays, dialogue coverage, shots that cut into a story — use `seedance-screenwriter` instead. That sibling skill outputs Seedance prompts in screenplay format and pairs with the `screenwriter` skill.
development
supadata
Extract transcripts and metadata from any video URL — Instagram, TikTok, YouTube, Twitter/X, Facebook — without opening a browser. Always try this FIRST before browser-use-cloud. Keywords: transcript, video metadata, supadata, transcribe video, youtube transcript, tiktok transcript.
devops
sync-plugin-skills
Sync skills from the laniameda-skills GitHub repo into the Cowork plugin. Use when Michael says "update the plugin", "sync the skills", "pull latest skills", "update plugin from GitHub", or "rebuild the plugin". This skill pulls the latest from github.com/Michailbul/laniameda-skills and rebuilds the laniameda AI creatorship .plugin file so Cowork picks up the changes.
tools
viral-psychology-hooks
Apply neuroscience-backed viral psychology principles to content ideation and storyboarding. Use when brainstorming ideas for Reels, carousels, short-form video, or any content that needs to capture attention and drive engagement. Generates concepts using morbid curiosity, gossip reward, prediction error, and Zeigarnik effect — the brain triggers behind million-view content. Triggers on: "ideation", "storyboard ideas", "viral concept", "content ideas", "hooks for", "make this scroll-stopping", "what would make this viral".
development
x-tweet-fetcher
Fetch tweets, replies, and user timelines from X/Twitter without login or API keys. Also supports Chinese platforms (Weibo, Bilibili, CSDN, WeChat). Includes camofox_search() for zero-cost Google search without API keys. Basic tweet fetching: zero dependencies. Replies/timelines/search: requires Camofox. NEW: X-Tracker for tweet growth monitoring with burst detection.
development
shader-landing-v0-reference
Recreate cinematic shader-first landing pages with full-screen horizontal sections, fixed generative backgrounds, translucent glass UI, custom cursor systems, magnetic buttons, and directional reveal choreography. Use when Codex needs to analyze a reference repo, screenshot, or live page and reproduce this specific design grammar or adapt these techniques in React, Next.js, or similar frontend stacks, especially when the reference is the v0 shaders landing page template.
development
scroll-storytelling
Build scroll-driven storytelling websites using AI-generated assets (images, videos, 3D models) with GSAP ScrollTrigger and Next.js. Combines AI creatorship workflow with scroll-hijacked cinematic web experiences. Use when: 'build a scroll site', 'scroll-driven page with AI assets', 'cinematic scroll experience', 'video scrub scroll', 'scroll storytelling', 'combine video and scroll', 'AI assets + scroll animation', 'build something like igloo.inc / Apple product page'. This skill focuses on ASSET INTEGRATION — how to wire AI-generated images, videos, and 3D models into scroll-driven timelines. For GSAP ScrollTrigger mechanics and patterns, defer to the scroll-experience skill. For Framer Motion scroll, defer to scroll-motion skill.
development
scroll-gsap-engine
GSAP ScrollTrigger master skill for building award-winning scroll-driven websites. Covers GSAP ScrollTrigger, Lenis smooth scroll, CSS scroll-driven animations, pinned sections, horizontal scroll, parallax, image sequences, text split reveals, and section transitions. Enforces performance rules, accessibility, and mobile-first architecture. Use when: GSAP, ScrollTrigger, Lenis, pinned scroll, scroll hijack, image sequence, SplitText, cinematic website, Apple-style scroll, dreamy scroll, scroll animation with GSAP. For React-only scroll with Framer Motion, use scroll-motion skill instead.
development
product-visual-generator
Generate production-ready AI product photography for any brand using Claude Code + Fal.ai (Nano Banana 2). Three-phase pipeline: scrape product images from a brand's website → analyze style references → generate styled product shots at 0.5K–4K resolution. Use when asked to create product visuals, ad creatives, product photography, or styled product images for a brand. Triggers on: "generate product visuals", "create product photography", "make product images for [brand/URL]", "style these product images", or when a brand website URL is provided in a visual/creative context.
development
parallel-web-search
DEFAULT for all research and web queries. Use for any lookup, research, investigation, or question needing current info. Fast and cost-effective. Only use parallel-deep-research if user explicitly requests 'deep' or 'exhaustive' research.
development
mac-disk-cleanup
Audit and clean up disk space on macOS. Scans all major space consumers — apps, caches, dev tools, audio plugins, sample libraries, node_modules, Docker, and system data. Produces a ranked report and executes removals on confirmation. Trigger when the user says "clean up space", "disk is full", "free up storage", "what's taking space", or "audit my disk".
tools
browser-use-cloud
Cloud browser automation for authenticated websites, dynamic pages, and multi-step web tasks. Use when supadata fails for video/social content, or when you need to log in, scroll carousels, fill forms, solve CAPTCHAs, or extract from pages that require a real browser session. Keywords: browser automation, scrape authenticated site, instagram scroll, extract dynamic page.
tools
visual-style-replicator
Deconstruct any visual aesthetic and turn it into a reusable generation system for branded images. Use when the user wants to steal, replicate, match, decode, systematize, or build prompts from a visual style such as vintage labels, posters, stamps, packaging, editorial design, or any recognizable art direction. Use this whenever the user shares style references and wants a repeatable prompt, master prompt template, brand system, or aesthetic breakdown — especially for Nano Banana Pro generation.
development
recraft-v4-prompting
Write high-performance prompts for Recraft V4 (V4 Standard, V4 Pro, V4 Vector, V4 Vector Pro). Use whenever the user asks for a Recraft prompt, a poster layout, editorial image, product shot, logo/icon/vector SVG, typographic design, or brand-ready visual to be generated in Recraft. Prefer this skill over generic image prompting when the target model is explicitly Recraft V4 or when the user needs print-ready, vector-native, or typography-integrated output.
development
nano-banana-pro
Generate/edit images and craft/critique/refine prompts for Nano Banana Pro / Nano Banana 2. Use for image create/modify requests, prompt optimization, structured prompt templates, JSON prompt formats, text-heavy layouts, typography, style transfer, color-grade transfer, locked-variable edits, and subject replacement. Supports text-to-image + image-to-image; 1K/2K/4K; use --input-image.
development
laniameda-pet-ad-pipeline
End-to-end AI pet ad generation. Use when the user wants to produce pet-focused ad clips — treat reactions, toy try-outs, product fits (collars/harnesses/beds/feeders), grooming reveals, day-in-the-life montages, or training moments. The pet is the hero, the owner's hands often in frame, voiceover optional. Orchestrates Nano Banana Pro (pet source image) + Seedance V2 (animation) + Claude Opus 4.6 (prompt optimization). Triggers on "make a pet ad", "generate pet UGC", "AI pet creator", "dog ad", "cat ad", "pet brand commercial", "BarkBox-style ad", "treat reaction video", or any request to turn a pet + product into ad-ready clips at scale.
tools
color-grade-transfer
Transfer cinematic color grading from a reference image onto a source photo using Nano Banana 2. Use when the user wants to apply a cinematic look, color style, mood, or tonal treatment from one image to another. Also triggers when user mentions: "match the colors", "apply this look", "transfer the grade", "make it look like this", "copy the mood", "apply color grading", "stylize my photo like this", "cinematic color transfer", or shares two images and asks to blend/match styles.
testing
laniameda-ugc-ad-pipeline
End-to-end AI-creator UGC ad generation. Use when the user wants to produce UGC-style ad clips with AI creators — reaction videos, virtual try-ons, unboxings, product holds, ad translations, or 3D template brand swaps. Orchestrates Nano Banana Pro (source image) + Seedance V2 (animation + lip sync) + Claude Opus 4.6 (prompt optimization). Triggers on "make a UGC ad", "generate UGC variations", "AI creator for this brand", "localize this ad", "A/B test this hook with N creators", or any request to turn a product + script into ad-ready UGC clips at scale.
tools
screenwriter
Скилл-инструмент для сценариста полнометражного фильма или сериала. Используй всегда, когда пользователь хочет писать сценарий, поэпизодник, разрабатывать сцены, бит-шит, диалоги, делать ревизии, считать экранное время, резать длину, работать с персонажами или мифологией истории. Скилл работает на основе методологий Макки, Кэмпбелла и Аристотеля, выдаёт Hollywood-формат .docx, поддерживает билингвальные сценарии (диалог на одном языке + перевод в скобках под ним), и помогает аудитировать структуру по причинности и движению ценности. Скилл не привязан к конкретной истории — пользователь приносит свою.
documentation
laniameda-gallery-query
Query the laniameda-gallery to browse, search, retrieve, and download vault content. Use when an agent needs to find assets, prompts, or structured designs-pillar references in the gallery and pull them into the current task.
data-ai
laniameda-gallery-ingest
Save prompts, images, tutorials, links, and design inspiration to the laniameda-gallery Convex knowledge base. Use when an agent needs to ingest prompts, files, URLs, or design references into the gallery and must stay aligned with the current repo ingest contract.
documentation
burst-frame
Generate BURST FRAME prompts for any subject — characters, creatures, products, environments, or objects — using the @kaigani technique in Seedance 2.0. Produces a 20-shot Rapid Fire prompt that outputs one video containing 20 consistent, full-resolution reference frames, then extracts them with ffmpeg. Use this skill whenever the user wants to: generate reference frames for a character, creature, or object, build a consistency sheet for any AI subject, get 20 angles/poses of anything, create a shot list for a non-automotive subject, or says "burst frame" for anything that isn't a car. Also trigger on: "reference frames for [subject]", "20 shots of [character/creature/product]", "consistency frames for [subject]", "burst frame [subject]", "[subject] reference sheet", "generate poses/angles of [subject]", "knight burst", "character burst", "creature reference", "product reference frames". For cars, use the burst-frame-cars skill instead — it has dedicated automotive angles.
development
seedance-screenwriter
Write Seedance 2.0 prompts in screenplay format for narrative storytelling — when the prompts will be cut into a film, short, or scene. Use whenever you're generating shots that will be edited into a continuous story with dialogue, character beats, scene continuity, or coverage. Pairs with the screenwriter skill — read the scene's screenplay first (or the project's `scene.md` if it exists), then translate each shot into a Seedance prompt that reads as a screenplay page, not as an engineering spec.
development
carousel-designer
Design and produce high-engagement social media carousels as self-contained HTML, rendered to PDF + PNGs via Playwright. Use when asked to build, generate, design, or iterate on a carousel, social media slides, Instagram carousel, LinkedIn carousel, slide deck, swipe post, or branded content slides. Supports Instagram, LinkedIn, Twitter/X, and Stories. Supports three modes: SHOWCASE (AI art / prompt shares), EDUCATE (tutorials, how-tos, listicles), and THINK (thought leadership / raw ideas). Also use for comparison carousels (tool-vs-tool, before-vs-after, model-vs-model) where each compared output should get its own full-width slide rather than a cramped split layout.
tools
brand-visuals-ai
Translates a brand identity into an AI visual production kit — a Visual DNA spec, prompt directions, and asset workflow for creating brand visuals with generative AI (text-to-image, image-to-image, image-to-video). Use this skill whenever the user is designing a brand and wants to create visuals with AI tools. Triggers include: "design visuals for my brand", "create AI prompts for my brand", "I want to design [product/brand] with AI", "help me make brand assets", "generate mockups for my product", "design a brand for X and create visuals", or any request combining brand work with generative AI image or video creation. Always use this skill BEFORE writing any image or video prompt for a brand project — it keeps everything consistent and on-brand. Works from a full brand brief OR a rough idea.
tools
desi-webdesign-practicies-extraction
Extract reusable web-design practices, interaction patterns, layout grammar, motion systems, and implementation techniques from a reference repo, component, code block, screenshot, or live URL. Use when Codex needs to study an existing design or frontend implementation and teach another agent how to recreate the same component, landing page, layout, or behavior in a different project, especially when complex patterns should be explained with source-backed code snippets rather than natural-language summary alone.
development
laniameda-x-post
Digest and learn from an X/Twitter post about AI creatorship, AI image/video generation, prompting, filmography, or creative workflows. Fetches the tweet, follows every linked resource autonomously, extracts all actionable techniques, saves to KB, optionally ingests to gallery, and encodes new knowledge into skills if warranted. Use when Michael pastes an X or Twitter URL about AI tools, prompting, Midjourney, FLUX, Kling, Runway, Seedance, ComfyUI, Stable Diffusion, AI filmmaking, or any AI creative workflow. Keywords: digest this tweet, learn from this tweet, extract from this post, what does this thread teach, ai prompts from tweet, follow this x link, upgrade skills from tweet, mine this, grab the prompts from this, digest this x post.
tools
ai-avatar-realistic
Expert prompting guide for generating photorealistic AI avatar images and video using Nanobanana 2 × Higgsfield (or similar tools like Midjourney, Runway, Kling). Use this skill whenever the user wants to generate or refine prompts for realistic AI avatars, creator/influencer setups, character reference sheets, product shots, or cinematic portrait-style images. Triggers include: "write a prompt for my AI avatar", "generate a realistic image prompt", "Higgsfield prompt", "creator room shot", "talking head prompt", "character reference sheet", "product influencer shot", "fix my skin texture prompt", or any request to produce or improve AI image generation prompts for realistic human subjects. Always use this skill when the user is building an AI influencer, AI avatar, or consistent character across multiple shots.
tools