qwwiwi

Audit and tune ANY existing site (static HTML, Astro, Next.js, WordPress export...) for search + AI discovery: SEO (Yandex & Google), AEO (AI answer engines), GEO (LLM/agent discovery). The agent checks robots.txt, sitemap.xml, canonical, meta/OG tags, JSON-LD, llms.txt and IndexNow, then generates whatever is missing. Use when the user says: optimize my site for search / SEO / чтобы находили в поиске / хочу в Яндекс и Google / AI-ответы. Tunes an existing site — it does not build one.

Generate Instagram carousels: 1080x1350 PNG slides from text (headline + body) and user photos, laid out in HTML/CSS and rendered headless via Playwright. Full design system baked in (Montserrat; cover/content/CTA card roles; calculated photo height; text-safe zones over faces; hyphen-only punctuation). Coral default palette, fully restylable to any brand (e.g. EdgeLab lime). Use WHENEVER you need a carousel for an Instagram post, want to lay out post slides, or turn a text + photos into a swipeable post. Triggers: "make a carousel", "carousel", "carousel post", "post slides", "slides for a post", "карусель", "слайды для поста", "пост-карусель", carousel/carousel. NOT for: single covers, reels, stories.

Generate premium, high-CTR YouTube thumbnails (1920x1080) — your real face from video frames, a punchy headline, a brand logo, on a dark high-contrast scene. Use this WHENEVER you need a YouTube cover/обложка/thumbnail, want a face pulled from a video onto a cover, want to "play with the font" on a cover, or want a thumbnail in a specific brand palette. Three engines: (A) GPT full-generation via gpt-image-2 with face+logo reference photos; (B) Compose — real rembg face cutout + scene + text via Playwright; (C) Hybrid — GPT no-text base + compose text for exact font control.

Research YouTube for content strategy — find what's taking off in a niche, scan competitor channels, rank videos by views and by view-velocity, and pull any video's transcript. Use this WHENEVER you need to research YouTube, find trending/viral videos on a topic, see what's blowing up on YT, scan a competitor's channel, get a YouTube video's transcript or 'what's in this video', or gather YouTube content ideas from what's working — even if the skill isn't named. Search and stats run keyless via yt-dlp; transcripts go through TranscriptAPI; a YouTube Data API key is optional for cleaner search.

Unified Telegram core (HTTP API + MCP stdio + tgdl + sales hooks) with single-session Telethon architecture. Use when setting up Telegram automation, reading/sending Telegram messages via local API, exporting chat history, or troubleshooting Telethon auth/session conflicts.

Self-audit, code quality improvement through clarifying questions, and proactive learning. Use when writing code, making decisions, completing tasks, or during weekly self-audit. Triggers on code review, task completion, weekly introspection, and when user says "интроспекция" or "аудит".

How to author, ship, and update content (lessons, bonuses) for the AgentOS intensive platform (<your-agentos-domain>). Covers source-of-truth pattern, mock layer, Supabase DB layer, block types, frontend rendering rules, deploy pipeline (build → tarball → prod server → PM2), and known pitfalls. Use when adding a new bonus, rewriting an existing lesson, changing block types, or shipping content updates to prod. Triggers: "добавь урок", "новый бонус", "обнови контент", "update lesson", "add bonus", "lesson 10 / 11 / 12", "intensive-agentos-platform", "<your-agentos-domain> content", "patch_db", "lessonN_steps". NOT for: backend logic changes, auth flows, payment integration, infrastructure work outside the content pipeline.

Archive messages from a Telegram business chat. Saves each message to daily markdown, classifies (DECISION/TASK/OPEN_QUESTION), makes daily summary, pushes decisions and summaries to storage. Use when: (1) a message is received from the monitored chat, (2) @mention of the monitor agent in this chat, (3) request for flush/day summary, (4) request for data from chat history of a partnership. NOT for: messages from other chats, direct user commands outside partnership context.

Fetch a SPECIFIC Twitter/X tweet, article, thread, profile, or timeline by URL/id via the SocialData API. Use when: a Twitter link, read a known tweet, check a profile, read a thread. For SEARCH / discovery / 'what's trending in a niche', use the `x-research` skill (Grok Live Search) instead — this skill is fetch-only.

Agentic skills framework (pcvelz fork): TDD, debugging, planning, code-review, git-worktrees, parallel agents. 15 skills with .tasks.json persistence between sessions.

Generate Excalidraw diagrams from JSON. Pipeline, mindmap, flowchart. Use when: draw diagram, visualize, architecture schema, mindmap.

Create charts, tables, and data visualizations via Datawrapper API

Web research via Perplexity Sonar API -- search, fact-check, analyze trends

Local kanban board for AI agents. Agents create/update/complete tasks, track status visually. Use when: manage tasks, kanban, create task, update status, task board.

Transcribe voice messages (.ogg) via Groq Whisper API. When you see <media:audio> in a message, extract the file path and transcribe it using the included script.

Google Workspace -- Gmail, Calendar, Drive, Sheets, Docs, Tasks. Use when: send email, check calendar, read/write spreadsheets, manage Drive files, Google Tasks.

Proactively diagnose an myfleet agent's gbrain MCP setup end-to-end: connectivity, identity, recall, swarm, hooks parity, webhooks, GitHub repo, MCP URL security, and skill install. Triggers: «gbrain-doctor», «gbrain health», «check gbrain», «gbrain MCP», «проверь gbrain», «диагностика gbrain», «не работает gbrain».

Browser automation CLI for AI agents. Use when the user needs to interact with websites, including navigating pages, filling forms, clicking buttons, taking screenshots, extracting data, testing web apps, or automating any browser task. Triggers include requests to "open a website", "fill out a form", "click a button", "take a screenshot", "scrape data from a page", "test this web app", "login to a site", "automate browser actions", or any task requiring programmatic web interaction.

Transcribe voice messages (.ogg) via Groq Whisper API. When you see <media:audio> in a message, extract the file path and transcribe it using the included script.

Instagram analytics via HikerAPI: top reels per account, watchlist of accounts to scan periodically, engagement scoring. Use when: (1) you need to analyze an Instagram account (followers, posts, top reels), (2) build a watchlist of competitors, (3) find the most-engaging reels of an account over a window of N days. Triggers: «analyze instagram account», «top reels», «watchlist», «engagement», Instagram profile URLs. NOT for: YouTube, Twitter, mass spam, auto-likes, auto-follows.

Clean Markdown extraction from any URL via markdown.new service. 80% token reduction for LLM use. Handles PDFs, images, audio, and embedded media. No signup required. Use when web_fetch fails or returns too much HTML noise. Triggers: "markdown.new", "clean markdown", "extract content", "convert to markdown"

Web research via Perplexity Sonar API with citations. Use for live web search, fact-checking, trend analysis, best-practice lookup. Triggers: «search», «research», «fact-check», «best practices», «perplexity».

Unified Telegram **user-account** core (Telethon HTTP API + helpers) for an agent that needs MTProto access. Use only when you actually need a user account (read history, write to chats where bots cannot, search messages, export chats). For ordinary bot integrations, do NOT use this — use the Telegram Bot API (`api.telegram.org/bot<token>/...`) directly.

Транскрипт YouTube роликов через TranscriptAPI.com. ОБЯЗАТЕЛЬНО используй этот скилл когда: (1) пользователь присылает YouTube ссылку (youtube.com, youtu.be), (2) просит рассказать о чём видео, (3) просит транскрипт/субтитры, (4) говорит «что в этом видео», «о чём ролик», «перескажи видео», (5) любой запрос связанный с содержимым YouTube видео. Триггеры: YouTube ссылка, «транскрипт», «о чём видео», «что говорят в видео», «перескажи ролик», «summarize video», «get transcript». НЕ использовать yt-dlp или youtube-transcript-api -- только этот скилл.

First-run personalisation wizard for a fresh EdgeLab agent. Invoke when the operator says "onboard me", "set up my agent", "personalise", or immediately after a clean install.

Turn a block of markdown or data into a self-contained, pretty HTML presentation and send it back over Telegram as a file. Invoke when operator says "present this", "make a slide deck", "visualise", "make an HTML report".

Schedule, list and cancel one-shot reminders via cron. Invoke when operator says "remind me at X", "in 10 minutes", "tomorrow at 9", "list my reminders", or "cancel reminder N".

Rewrite or tune the agent CLAUDE.md + rules.md + USER.md based on accumulated LEARNINGS.md and operator feedback. Invoke when the operator says "recompile yourself", "update your rules", or when LEARNINGS.md has grown beyond ~30 entries.

Двойное ревью: вторая модель проверяет работу первой. Используй когда: (1) закончен значимый кусок кода перед коммитом или PR, (2) владелец просит «прогони кросс-ревью» / «двойное ревью», (3) впереди деплой, оплата или другой необратимый шаг. Правило курса AgentOS: одна модель проверяет себя плохо -- свежий взгляд ловит то, что автор не видит.

Post-install onboarding flow. Agent greets the user, activates self-compiler, and guides them through configuring CLAUDE.md and workspace files through natural conversation. Use after fresh Claude Code install. Covers identity, goals, style, stack, and generates all config files. The whole flow takes 5-10 minutes.

Generate beautifully formatted standalone HTML documents with clean minimal design. Triggers: /present, 'make a presentation', 'visualize', 'format as HTML', 'create a document'. Accepts any content (lesson, report, comparison, roadmap, dashboard) and produces a single HTML file with light/dark theme toggle. Zero external dependencies.

Compile human thoughts, questions, and conversations into structured knowledge files for the agent. The agent learns who you are and becomes smarter with every conversation.

Создание новых элементов воронки в ChatPlace: кодовые слова с DM-автоматизацией и публичным ответом под комментом, AI-агент-продажник с knowledge base + topic rules + global rules. Use when: (1) adding new Instagram keyword that must trigger DM + public comment reply, (2) configuring or updating the ChatPlace AI sales agent (your account), (3) writing/editing global rules or topic rules for the AI agent, (4) adding Q&A entries to knowledge base, (5) testing agent answers before publishing, (6) troubleshooting ChatPlace setup (API limits, AI Pack subscription errors). NOT for analytics/reports (out of scope — keep in your own tracking system).

Полный инструмент для работы с Instagram: аналитика аккаунтов через HikerAPI SaaS + скачивание рилсов/постов через Cobalt API. Используй когда: (1) скачать рилс/видео/фото из Instagram, (2) найти залетевшие рилсы аккаунта, (3) проанализировать конкурента (views, likes, comments, engagement), (4) управлять watchlist аккаунтов для мониторинга, (5) скачать + транскрибировать видео, (6) поиск по хэштегам, локациям. Триггеры: «скачай рилс», «скачай из инсты», «проанализируй аккаунт», «залетевшие рилсы», «топ рилсы», «watchlist», «отследить аккаунт», «конкуренты инста», «engagement», ссылки на instagram.com/reel/, instagram.com/p/, instagram.com/stories/. НЕ для: YouTube, Twitter, массового спама, автолайков, автоподписок.

Use when a brand, company or AI-tool logo is needed for a banner, cover, post or composite (найди лого X, иконка Y, логотип Z, add brand icon, логотип компании) — Claude, ChatGPT, Codex, OpenAI, Anthropic, Cursor, Telegram, GitHub, Notion, Gemini, Perplexity, Midjourney or any other brand. Looks up the official SVG in a local library (glincker/thesvg, 5886 brands, MIT) and returns a file path (default or mono variant). NOT for: photos of people, custom illustrations, AI-art — official brand SVGs only.

Use when asked to generate an image, picture, photo, cover, banner art or illustration (сгенерируй картинку, фоточку, обложку, изображение) using the Codex/ChatGPT subscription — no OPENAI_API_KEY, no per-image billing; supports up to 5 reference images (по референсу, в стиле картинки). Also use when tempted to answer that image generation is impossible without an API key, that Codex is text-only, or that references don't work over the subscription.

Senior full-stack architect for educational SaaS platforms (Teachable/Thinkific/Kajabi level). Provides decision frameworks, stage-aware stack recommendations, architecture patterns, AI/LLM integration, agent-native design, security threat modeling, testing strategies, and MCP integration guidance. Triggers: /senior-brainstorm, "brainstorm", "architecture", "how to build", "stack selection", "tech choice", "architectural decision", "design this", "senior brainstorm", "platform stack". NOT for: simple CRUD tasks, non-EdTech projects, single-framework questions without product context.

Build production-grade Telegram bots from scratch -- architecture, code, database, deployment, monetization. Supports Python (aiogram 3.x, python-telegram-bot v21) and TypeScript (grammY). Database flexibility: SQLite local, Supabase (PostgREST + Edge Functions + RLS), PostgreSQL (asyncpg, Prisma, Drizzle), Yandex Cloud, Redis. Covers webhook/polling, FSM, middleware, AI integration (OpenAI/Claude streaming), Telegram Stars payments, Mini Apps, Business API. Use when: (1) creating a new Telegram bot, (2) adding features to existing bot, (3) choosing stack/database for a bot, (4) deploying a bot to production, (5) integrating AI/LLM into a bot, (6) adding payments/Stars, (7) scaling a bot, (8) securing webhook endpoints. Triggers: telegram bot, tg bot, aiogram, grammy, bot api, telegram payments, telegram stars, bot webhook, bot polling, bot fsm, bot deployment, inline keyboard, conversation handler, mini app bot.

Use when an EdgeLab Space member queries the community library — skills, lessons, usecases, lives, workshops, guides — or reads hub questions or the community digest from the CLI/terminal with curl, no MCP client. Triggers — найти навык в EdgeLab Space, поискать урок или разбор, открыть полный материал по slug, забрать копируемый промпт скилла, посмотреть дайджест сообщества за период через curl. For MCP-native clients (Claude Code, Codex, cowork, Cursor) prefer adding the server with a header — see docs/MCP.md.

Use when an EdgeLab Space member's agent needs to participate in AgentTinder — the community Q&A hub on platform.edgelab.space. Triggers — задать вопрос от агента, ответить на вопрос в AgentTinder, проверить ленту вопросов, оставить комментарий, поставить лайк, отметить решение, прочитать ленту сообщества через агентский API. Covers the full auth flow (ea_ key → identity JWT) and every /via-agent endpoint, karma, limits and etiquette. NOT for searching the knowledge library (use edgelab-knowledge-search for that).

Diagnose and safely complete a migration from the legacy Telegram gateway to the dashi-plugin-claude-code channel. Use this whenever an agent is moving off the old gateway, when a freshly migrated bot is silent / hangs / answers as "default Claude", when you hit a Telegram 409 conflict, or when planning the cutover. Trigger on phrases like "переезд на плагин", "migrate from gateway", "доктор для переезда", "бот молчит после миграции", "409 conflict", "channel plugin not responding". Run the doctor BEFORE and AFTER cutover — it encodes every mistake we already paid for so a student's agent migrates without repeating them.

Senior full-stack architect for educational SaaS platforms (Teachable/Thinkific/Kajabi level). Provides decision frameworks, stage-aware stack recommendations, architecture patterns, AI/LLM integration, agent-native design, security threat modeling, testing strategies, and MCP integration guidance. Triggers: /senior-brainstorm, "brainstorm", "architecture", "how to build", "stack selection", "tech choice", "architectural decision", "design this", "senior brainstorm", "platform stack". NOT for: simple CRUD tasks, non-EdTech projects, single-framework questions without product context.

Запуск онлайн-продукта по методу Джеффа Уокера – Product Launch Formula (PLF). Используй когда нужно: (1) спланировать запуск курса/инфопродукта/коучинга/SaaS с лимитом мест, (2) построить таймлайн от даты Open Cart назад через Pre-Launch и Pre-Pre-Launch, (3) написать серию подогревающих постов или писем перед стартом продаж, (4) сделать структуру трёх PLC-видео (Opportunity / Transformation / Ownership), (5) спроектировать ежедневную последовательность для Open Cart периода с финальным crescendo, (6) применить ментальные триггеры Уокера (scarcity, social proof, anticipation, reciprocity, community, authority, events), (7) адаптировать PLF под Telegram-канал, email-лист или вебинар. Триггеры: «запуск», «launch», «PLF», «Walker», «Уокер», «open cart», «pre-launch», «product launch formula», «продуктовый запуск», «как запустить курс», «sideways sales letter», «PLC1», «PLC2», «PLC3», «контент-план запуска», «как разогревать перед запуском», «когда открывать продажи», «как продать с лимитом мест».