Web & Frontend

serejaris Bundle

macOS / Windows 只读存储分析助手（自动识别系统）。扫描整机磁盘占用，找出占空间大户，把每一项分成 🟢可自动清理 / 🟡需人工判断 / 🔴谨慎清理 三级并给出可执行处置方案，生成排版精美、可折叠、命令可一键复制的交互式 HTML 报告，并可起本地服务在网页上一键删除（移废纸篓/直接删）。扫描全程只读。务必在以下场景使用：用户说"存储分析"、"磁盘满了"、"C盘/硬盘满了"、"空间不够"、"清理空间"、"清理磁盘"、"占空间"、"哪些东西占地方"、"帮我看看存储"、"看一下电脑存储/空间"、"存储空间"、"电脑空间不够"、"内存满了/不够/不足"、"看下内存/存储"（中文口语里"内存"常指存储空间）、"storage analysis"、"disk cleanup"、"清缓存"、"磁盘清理"，或用户抱怨电脑没空间、想知道什么东西吃硬盘、想要清理建议时。注意：若用户明确指运行内存/RAM（如"哪个进程吃内存""内存占用高"想看活动监视器），那是 RAM 不是存储，不属于本 skill。

serejaris Bundle

Build any web-based project: websites, landing pages, web apps, dashboards, browser games, portfolios, and interactive experiences. Design-first React workflow. Skip if the user explicitly requests a non-React framework (Vue, Svelte, Angular, vanilla HTML) or the task is unrelated to web UI (CLI tools, scripts, data pipelines).

runxhq Bundle

Extract schema-validated JSON from messy HTML or text fixtures with digest-bound provenance.

saeed-vayghan Bundle

Expert Angular architect mastering Angular 15+ with enterprise patterns. Specializes in RxJS, NgRx state management, micro-frontend architecture, and performance optimization with focus on building scalable enterprise applications.

saeed-vayghan Bundle

Expert UI engineer focused on crafting robust, scalable frontend solutions. Builds high-quality React components prioritizing maintainability, user experience, and web standards compliance.

saeed-vayghan Bundle

Expert accessibility tester specializing in WCAG compliance, inclusive design, and universal access. Masters screen reader compatibility, keyboard navigation, and assistive technology integration with focus on creating barrier-free digital experiences.

ever-just Skill

Remotion — Programmatic Video Creation with React

ever-just Skill

Motion (Framer Motion) — React UI Animation Library

ever-just Bundle

Evaluate user experience by locating decisions on a layered framework drawn from Jesse James Garrett's Elements of User Experience and modern frontend practice. Use this skill when reviewing or auditing any product interface, when diagnosing a UX problem (where does it actually live?), when designing or systematizing a frontend, when discussing information architecture, interaction design, layout, design tokens, or design systems, when asked questions like "is this a UI problem or a deeper problem", "what should we systematize", "where should this decision be made", or any request to evaluate, critique, or improve a digital product holistically. Trigger this skill whenever a question touches the structure of a product rather than just its surface — even if the user doesn't say "UX" explicitly.

ever-just Skill

Operate an everjust.app (Odoo 19) tenant from the BOX SHELL (SSH + docker), the deploy-pipeline and box-ops layer. Use when the task needs bulk DB writes, DB-only page publishing, clearing the sitemap ir.attachment cache, creating an ir.cron, recovering a CI deploy whose rsync silently did not land, avoiding deploy-collisions across concurrent agents (gh run idle check, scoped git add), restarting Odoo to compile a QWeb edit, or fixing an nginx stale-inode after a config redeploy. Covers the odoo shell invocation, the COW-fork-overrides-generic gotcha, and the Deploy to Production pipeline. The deploy/ops layer around [[everjust-website-customization]] (which owns durable view/CSS/behavior customization from the shell). Prefer the MCP ([[everjust-agent-mcp]], [[everjust-website]]) for ordinary edits; drop to the shell when it cannot. Cross-references [[everjust-platform]], [[github-actions-ec2-deploy]].

ever-just Bundle

Orchestrate many parallel subagents to run a large multi-file refactor, migration, or codemod sweep SAFELY and fast — partition by disjoint file ownership, fan out one write-agent per independent slice behind typecheck+build barriers, and synthesize their structured returns. Use when adopting a component library across routes, migrating design tokens / dark mode, running a rename or codemod sweep, modernizing per-route, or applying one rubric across a whole codebase — any breadth change that decomposes into independent file sets. Not for interdependent foundation work (shared config, a token layer, coupled modules) — do that inline.

ever-just Bundle

Add dark mode to a light-only app that styles with RAW color utilities (Tailwind `stone-*`/`slate-*`/`gray-*`) via a semantic token layer, migrating utilities onto it so light stays PIXEL-IDENTICAL and dark comes for free. Use when asked to "add dark mode", or to refactor hardcoded color utilities into a themeable token system. Not for apps already on semantic tokens — just add `.dark` values plus a toggle.

ever-just Bundle

Audit a web UI for mobile/layout defects EMPIRICALLY by rendering every servable surface across a phone/tablet/desktop viewport matrix (both orientations, light + dark) with a Playwright DOM probe that MECHANICALLY flags horizontal overflow, sub-44px tap targets and sub-16px inputs, then ships the probe as a non-zero-exit pre-merge gate. Use when the frontend "isn't optimized for mobile", text "spills outside its container", small-screen stacking is "awkward", or for any responsive / tap-target / layout audit or checked-in gate. Not for pure visual-taste questions — use `ui-ux-audit` / `frontend-design`.

ever-just Bundle

Scaffold a token-based shadcn/ui + Radix primitive set (select/combobox/command/switch/tooltip/form/drawer/…) matching a codebase's house conventions on Tailwind v4 — semantic-token colors, a global focus ring, no tailwindcss-animate, 44px touch targets, forwardRef like the existing button. Use when an app hand-rolls or fakes controls (native `<select>`s, `title=` tooltips, custom dropdowns, DIY meters), or when standing up shadcn/ui on Tailwind v4.

ever-just Skill

Deeply and durably customize an EverJust.app (Odoo) tenant website from the odoo shell — the layer below the MCP website_* tools. Use when a change needs QWeb view edits, site-wide CSS/JS, new server-side behavior, or model/config writes that the everjust MCP and website builder cannot do (editing views, server actions, custom_code_head, base automations). Covers COW inherit views (website_id=1) that survive addon redeploys, the website.custom_code_head Markup-escaping trap, base_automation for behavior, the "view edits need docker restart to render" rule, and the durable-vs-reverts-on-upgrade matrix. Use when the user says "edit the Odoo view/template", "inject CSS/JS site-wide", "add a server action / automation", "the change keeps getting reverted on deploy", or "customize the everjust/connectdomain website beyond the MCP tools".

ever-just Bundle

Convert markdown files to styled PDFs using pandoc and WeasyPrint with CSS. Use when exporting markdown documents to PDF with custom styling.

nirholas Skill

Guide to Bitcoin inscriptions (Ordinals) — creating, minting, and managing content inscribed on the Bitcoin blockchain. Covers Ordinal theory, inscription types (text, images, HTML), BRC-20 tokens, recursive inscriptions, and marketplace integration.

nirholas Skill

Guide to W3AG (Web3 Accessibility Guidelines) — accessibility standards and patterns for Web3 applications. Covers wallet connect UX, transaction confirmation patterns, error handling, screen reader support, and inclusive design for dApps. Making DeFi usable by everyone.

nirholas Skill

Guide to building a real-time crypto market dashboard with Next.js, React, and TradingView charts. Features multi-chain portfolio tracking, DeFi position monitoring, price alerts, and social sentiment integration. Production-ready dashboard template.

gked2121 Skill

Convert a light-only site or app to a real dark mode -- semantic token mapping, elevation-based surfaces, recalibrated shadows and imagery -- not a naive inversion. Handles Tailwind, CSS variables, and styled-components, plus the toggle, persistence, and no-flash boot.

gked2121 Skill

Audit a website or UI and de-AI it — find the patterns that make it look template-generated (in both design and copy) and return specific, prioritized fixes that push it toward editorial, premium, human-crafted. Reviews layout, color, type, spacing, motion, accessibility, AND the words. Use after building or generating any frontend, or when a site looks "fine but generic" and you can't say why.

gked2121 Bundle

Create professional CSS animations, transitions, micro-interactions, and complex motion design. Use when adding animations, hover effects, loading states, page transitions, scroll animations, or any motion design work.

gked2121 Skill

Install a complete design-token theme into any project in one pass -- pulls a style from the Open Agent Stack design-styles collection (aurora-mesh, cirrus, liquid-glass, mono-brutalist, neo-terminal, sand-terra, tidal) or a local tokens.json, wires it into Tailwind/CSS, and repaints existing components to match.

gked2121 Bundle

Create error boundaries, error handling, and fallback UIs for React applications. Use when implementing error handling, creating fallback components, or setting up error reporting.

gked2121 Bundle

Uses 1M context window to ingest an entire codebase and output a file-by-file migration plan. Supports JS to TS, React class to hooks, framework migrations, and more. Generates migration-plan.md with file inventory, dependency graph, migration order, file-by-file changes, estimated effort, and risk assessment.

gked2121 Bundle

Uses Managed Agents' 14.5-hour runtime to audit an entire codebase overnight. Security, performance, accessibility, dependency issues. You wake up to a full report.

gked2121 Bundle

Create design systems with tokens, components, and documentation. Use when building design systems, creating component libraries, or establishing design tokens.

agentik-os Skill

Publishes deliverable reports (audit, research memo, strategy doc, mission recap, dashboard, brief) as LIVE claude.ai Artifacts via the native Artifact tool; surface 1 of the OmegaOS report router (R-ARTIFACT). Use when the user says "report", "live report", "publish an artifact", "make this a report", "dashboard report", "share a page", or in French "rapport", "rapport live", "publie un artifact", "fais-moi un rapport", "tableau de bord". NOT for explicit PDF asks (omega pdf, R-PDF), plain file-only asks (R-HTML), generating videos or images, or scripted browser E2E.

agentik-os Bundle

Claude Changelog Adopt — a daily autonomous self-improvement pass that reads the OFFICIAL Claude Code changelog (the authoritative GitHub CHANGELOG for anthropics/claude-code, with @claudecodelog as a secondary mirror), diffs it against the last version OmegaOS already absorbed, and for each NEW entry decides whether it implies a concrete OmegaOS or system-agent improvement — a new primitive, a doctrine change, an agent-behavior shift, a deprecation to unwind. It classifies each entry (opus), proposes a surgical adoption (what to change, in which file, why — like the manual /loop -> R-LOOP native-loop doctrine move), runs an adversarial anti-fabrication gate (opus, default-reject), writes a self-contained HTML report + Telegram alert for the operator to decide, and — only when ARMED — dispatches each vetted proposal to an oracle behind the full quality gate (In-Review handoff, never auto-Done, never force-push). Auto-adoption is scoped to doctrine (rules.rs), agent identities (agents/*.md), skills, and instal

agentik-os Skill

Forensic accessibility audit v1 (Gestalt-Popper). 21-phase deep analysis of every accessibility surface: WCAG 2.1 AA compliance, keyboard navigation (every interactive element), screen reader testing, ARIA labels/roles/states, color contrast (4.5:1 AA, 3:1 large text), focus management, skip navigation, form labels, error announcements, alt text, heading hierarchy, landmark regions, touch targets (44px), motion/animation preferences, cognitive load, reading level, plus verdict, fix plan, fix execution, re-audit, and integration smoke gate. Score /320. Preamble v1.0 compliant. Audit -> Plan -> Fix -> Re-audit. Use when user says "/a11yaudit", "accessibility audit", "wcag audit", "a11y check", "keyboard navigation test", "screen reader test", "color contrast check".

agentik-os Skill

Forensic user flow audit v2. 25-phase deep analysis of every user journey, workflow, state machine, edge case, error recovery, and experience gap. Covers flow mapping, state verification, dead-end detection, permission gaps, data integrity through flows, onboarding completeness, cross-session continuity, error recovery paths, accessibility of journeys, flow performance, plus verdict, fix plan, fix execution, re-audit, and integration smoke gate. Score /400. Preamble v1.0 compliant. Audit → Plan → Fix → Re-audit. Use when user says "/flowaudit", "audit the flows", "check user journeys", "workflow audit", "flow review".

agentik-os Skill

Forensic internationalization & localization audit v1 (Gestalt-Popper). 18-phase deep analysis of everything that decides whether the product is WORLD-READY: hardcoded user-facing string detection (JSX text, alt/title/aria/placeholder attributes, toast/error/email/notification copy), i18n framework wiring integrity (key existence, namespace structure, interpolation contract), locale routing & detection (path/subdomain/cookie/Accept-Language, persistence across navigation), pluralization rules (CLDR plural categories: zero/one/two/few/many/other), gender/grammatical agreement, date/time formatting (Intl.DateTimeFormat, timezone), number/percent/unit formatting, currency formatting (ISO 4217, symbol placement, minor-unit precision), RTL support (dir attribute, logical CSS properties, bidi-safe interpolation, mirrored icons), translation completeness + fallback chains, character encoding (UTF-8 end to end, normalization, mojibake), locale-aware sorting & collation (Intl.Collator), and untranslated-UI leakage at

agentik-os Skill

Forensic motion design audit v1 (Gestalt-Popper). 23-phase deep analysis of everything that MOVES on the site: CSS transitions, JS animations, WebGL effects, scroll-driven choreography, P5.js/canvas, page transitions, micro-interactions, loading sequences, easing systems, duration consistency, choreography composition, reduced-motion compliance, mobile motion, brand motion DNA, performance budget, plus verdict, fix plan, fix execution, re-audit, and integration smoke gate. Score /360. Preamble v1.0 compliant. ABORTS on non-UI projects (CLI, library, backend-only, headless). Audit -> Plan -> Fix -> Re-audit. Use when user says "/motionaudit", "motion audit", "animation audit", "why does it feel lifeless", "motion review", "animation quality", "easing audit", "scroll animation check", "does the motion make sense", "audit the animations".

agentik-os Skill

Senior Dashboard REFONTE engine v2. Three modes: EVOLUTION (default — surgical improvement), REVOLUTION (ground-up when unsalvageable), VERIFY (post-impl measurement 2-4 weeks later). 25-phase pipeline ALWAYS, scored /540, driven by real user data + user stories. Keep Audit classifies KEEP/IMPROVE/RETHINK/KILL. Component Forge writes real shadcn code. Reference patterns from shadcn official blocks/components (Context7 MCP) + product design knowledge. Confidence score per proposal. shadcn/ui + Tailwind v4 foundation. Use when user says "/refontaudit", "/refonte", "refonte", "refund audit", "redesign dashboard", "repenser dashboard", "comme Linear", "comme Vercel", "comme ElevenLabs", "dashboard pro", "dashboard senior", "vérifie la refonte", "verify refonte".

agentik-os Skill

Senior UI/UX Engineer. Architect digital interfaces overriding default LLM biases. Enforces metric-based rules, strict component architecture, CSS hardware acceleration, and balanced design engineering.

agentik-os Skill

Intelligent audit orchestrator — detects project type + user intent, recommends optimal audits with three bounded modes (Focused/Standard/Forensic). Use when user says "/audit", "what should I audit", "full audit", "audit my project", "audit fast", "audit deep", "find issues", "improve quality", "production ready check", "ship-ready audit". Auto-detects project stack and intent keywords (speed, security, design, content, accessibility, full) to pick best 1-N audits. Dispatches in parallel waves. Reads results from audits/.<audit-id>/verdict.json after each run.