Claude Skills by legioncodeinc

Runs Lighthouse and PageSpeed Insights for React/Next.js: CI setup (LHCI), audit categories, performance budgets, lab-vs-CrUX gap. Use when setting up or debugging Lighthouse CI.

Customer support surface specialist — Intercom, Crisp, Plain, Pylon, Help Scout — widget integration, HMAC/JWT identity verification, conversation routing, AI deflection (Fin 2.0, Ari, Crisp Bot), and the data-export discipline. Use when the user says "integrate live chat", "set up Intercom", "add a support widget", "wire HMAC identity verification", "configure AI deflection", "design conversation routing", or "set up customer support for our SaaS". DO NOT use for managing deployments (devops...

Audit Lovable-built Supabase apps: RLS denials, role-differential reads, function gating, auth posture, bundle-forensics key/endpoint extraction, findings triage, reporting.

Markdown/MDX processing specialist - MDX 3/compiler selection, remark/rehype plugin pipelines, Shiki v4/expressive-code/starry-night syntax highlighting, GFM, AST manipulation, custom directive plugins, math/Mermaid diagram embedding, and XSS sanitization. Use when building or auditing any content processing pipeline that takes .md/.mdx source to HTML/JSX output.

MCP protocol authority for The Hive - builds and audits MCP servers and tool contracts against the Model Context Protocol spec and @modelcontextprotocol/sdk. Covers tool vs resource vs prompt design, zod input schemas (incl. the zod v3/v4 SDK trap), stdio vs HTTP transport, JSON-RPC framing and error semantics, capability negotiation, auth for remote/HTTP servers (API keys, bearer, OAuth 2.1), testing, and registering a server in Claude Code, Cursor, Codex, and Cowork. Includes a worked examp...

Documents MCP tools, TypeScript APIs, and CLIs honestly: schema, output, side effects, examples, doc-to-code sync. Use when writing or auditing tool/API/CLI docs.

Reviews, audits, and extends the AI cognitive layer: coach routing, prompt cascade, RAG, three-tier memory, observability, evaluation. Use when reviewing AI code or tuning retrieval.

Accessible overlay specialist for React: primitive selection (Radix Dialog, AlertDialog, Vaul Drawer, Sonner toast, cmdk command menu, Headless UI), the six-point accessible-modal contract (focus trap, escape, scroll lock, aria-modal, aria-labelledby, focus return), and the four-tier toast-vs-notification taxonomy. Use when choosing an overlay primitive, debugging focus trap regressions, implementing a drawer/sheet, wiring Sonner toasts into a React app, building a command palette, or auditin...

Photographic generation and editing for consented model work using gpt-image-2 or Nano Banana Pro. Supports capture-backed references and explicitly acknowledged visual-only references, with capability-gated editing, imperfection-led realism prompting, and honest EXIF handling.

Neon Postgres and Drizzle ORM on SvelteKit/Vercel: connections, schema, migrations, branching, query patterns, pgvector, RLS, and migrating off Supabase. Use for any Neon or Drizzle task.

Newsletter-as-channel specialist for product builders — Beehiiv, ConvertKit (Kit), Loops, Substack, Resend Audiences, Ghost, the embedded-newsletter-page pattern, deliverability vs platform tradeoffs, and monetization options (ad network, paid subscriptions, sponsorships, referrals). Use when the user says "which newsletter platform should I use", "set up Beehiiv", "embed a newsletter signup", "migrate from Substack to Beehiiv", "how do I monetize my newsletter", "Loops vs Kit for my SaaS", "...

OKR methodology specialist for writing, grading, and iterating on Objectives and Key Results. Enforces the output-vs-input discipline, diagnoses sandbagged vs. ambitious goal-setting, calibrates quarterly cadence and check-in rituals, contextualizes OKRs against KPIs and MBOs, and adapts the framework for small teams and startups. Activate when the user says "write OKRs", "audit our OKRs", "are these KRs measurable?", "set up a quarterly goal cycle", "OKR vs KPI", "OKR for small team", "grade...

Stripe for SvelteKit. Defaults to custom checkout with Elements (Payment/Address/Express Checkout, Appearance API) over hosted Checkout. Covers Payment Intents, subscriptions, webhooks, PCI scope.

PostHog for SvelteKit: install, pageview tracking, autocapture vs manual events, identify/alias, feature flags, experiments, replay, surveys, groups, Vercel proxy, EU/US residency, cost control.

Preact 11 specialist skill: signals-based reactivity (v2 API with createModel/useModel), preact/compat migration from React, third-party embed widgets (shadow DOM isolation, IIFE bundles), Astro island integration (client:* directives, 5.0.1 or newer useId fix), and Fresh 2.x framework (Deno-native, islands + serializable props). Use when building Preact components, evaluating Preact vs React, migrating from React to Preact, embedding a widget on third-party pages, or working with Astro or Fr...

Customer-feedback-to-roadmap loop specialist — Userback, Canny, Featurebase, Productboard, Frill, Productlane — platform selection decision tree, in-app-widget vs portal vs voting-board taxonomy, de-duplication discipline, RICE/ICE prioritization, status-transition policy, public vs private roadmap playbook, and CRM/issue-tracker integration wiring. Use when the user says "set up a feedback system", "which feedback tool should I use", "Canny vs Featurebase", "our feature requests are a mess",...

In-app product tour and onboarding UI specialist. Selects the right tour tool (Userpilot, Appcues, Userflow, Pendo Guides, Driver.js, Shepherd.js, Intro.js), implements tooltip/modal/hotspot/checklist components, wires segment-based trigger logic, and establishes a tour maintenance protocol that survives iterative UI changes. Invoke when the user says \\\\\\\"set up a product tour\\\\\\\", \\\\\\\"build an onboarding checklist\\\\\\\", \\\\\\\"compare Driver.js vs Shepherd.js\\\\\\\", \\\\\\\...

Python code specialist for Django, FastAPI, Celery, Channels, pytest, uv, and Ruff. Use when reviewing, refactoring, or authoring Python code, or migrating DRF to Ninja.

Audits a completed implementation against its source plan document and produces a structured findings report. The report goes in the source plan's `reports/` subfolder (e.g., `library/requirements/{lifecycle}/prd-{###}-{title}/reports/{date}-qa-report.md` or `library/issues/{lifecycle}/ird-{###}-{title}/reports/{date}-qa-report.md`); standalone audits go to `library/requirements/reports/{domain}/{date}-qa-report.md`. Use when the user says \\\\\\\"QA this\\\\\\\", \\\\\\\"audit the implementa...

Forge new rules, plugins, commands, agents (Bees), and skills (Stingers) for The Hive across Claude Code, Cursor, Codex, and Cowork. Use when creating, updating, or validating any Hive component.

Placeholder stinger shipped inside the Hive reference plugin skeleton, showing where a real plugin-bundled skill lives. Not a working stinger on its own.

{One or two sentences: what this stinger does and exactly when Claude should reach for it. Front-load the trigger words, since Codex and Cursor both judge relevance from this text alone. Keep it under 200 characters if you want the same file to survive Cowork's stricter cap without editing.}

Reviews, refactors, and authors React 18/19 codebases using the bulletproof-react architectural pillars and the curated 2025-2026 React ecosystem. Use when the user says \\\\\\\"review this React code\\\\\\\", \\\\\\\"react architecture review\\\\\\\", \\\\\\\"audit our React app\\\\\\\", \\\\\\\"state management decision\\\\\\\", \\\\\\\"what's the bleeding-edge React pattern for X\\\\\\\", \\\\\\\"propose a React refactor\\\\\\\", \\\\\\\"is this React anti-pattern\\\\\\\", or when `react-w...

React-to-Svelte 5 porting: hooks-to-runes translation, TanStack Router to SvelteKit, RTK Query and Radix conversion, contract-first port waves. Use when porting React components or dashboards.

Authors, audits, and restructures README files so they convert visitors into users. Apply when the user says \\\\\\\"write a README\\\\\\\", \\\\\\\"audit my README\\\\\\\", \\\\\\\"make my README better\\\\\\\", \\\\\\\"README for this project\\\\\\\", \\\\\\\"README-driven development\\\\\\\", or when starting a new project and the README does not exist yet. Also apply when badges are broken or missing, the quickstart is not copy-paste runnable, or the user wants to differentiate between an...

Designs and audits retrieval: full-text search, pgvector recall, hybrid RRF fusion, reranking, chunking, recall quality eval. Use when tuning recall or a search query misses.

Equips retrospective-worker-bee to run retrospectives that actually change behavior: format selection (Start/Stop/Continue, 4Ls, sailboat, mad/sad/glad, DAKI, Starfish, and more), psychological safety pre-check (Edmondson scale), facilitation playbooks, async retro design, and action-item follow-through discipline. Use when the user says \\\\\\\"run a retro\\\\\\\", \\\\\\\"plan our retrospective\\\\\\\", \\\\\\\"which retro format should we use\\\\\\\", \\\\\\\"our retros produce no change\\...

Review collection and online-reputation specialist for SaaS products -- G2, Capterra (now G2-owned), Trustpilot, Product Hunt, AppSumo, and Software Advice. Covers platform selection, profile setup, in-product review-request UX (timing, two-step pattern, copy), G2 incentive compliance (2026 rules + FTC Consumer Reviews Rule), Product Hunt launch-day playbook (00:01 PT, first-6-hours intensity, hunter sourcing), negative-review response templates, and earned-badge-as-conversion-asset deploymen...

Operational runbook authorship specialist covering canonical templates (break-fix, scheduled operation, diagnostic), the no-implied-context audit protocol, exact-command discipline, escalation path architecture, rollback procedure standards, runbook-as-test (game day) methodology, and postmortem-to-runbook linkage. Activate when the user says \\\\\\\"write a runbook\\\\\\\", \\\\\\\"audit this runbook\\\\\\\", \\\\\\\"our runbooks are out of date\\\\\\\", \\\\\\\"we need a runbook for this al...

Rust development specialist for Cargo, async services, persistence, CLI/TUI, tests, upgrades, and current toolchains. Use for Rust implementation, review, migration, or release guidance.

Security audit for SvelteKit, Neon/Drizzle, WorkOS, Stripe, Vercel, Doppler, and GoHighLevel. First gate of the Ship Gate - scans, triages, and remediates vulnerabilities before quality-stinger.

Sentry error tracking for SvelteKit on Vercel - client/server hooks, source maps and releases, tracing sampling, session replay, PII scrubbing, alert tuning, and cost control.

SEO and AEO for SvelteKit (Svelte 5) + Payload CMS + Vercel. Covers metadata, sitemap/robots, JSON-LD, Payload SEO fields, Core Web Vitals, llms.txt/AI citation, E-E-A-T, and indexation playbooks.

The shadcn-svelte library itself for any Svelte project: CLI, copy-in model, anatomy, registry, theming, forms, upgrades. Not OSPRY's design system on portal/web/wl (ux-ui-svelte-stinger owns that).

Slack app development specialist for Bolt SDK (JS/Python/Java), slash commands, Block Kit UI composition, modals, Events API, OAuth multi-workspace install, and App Directory submission. Use when the user says \\\\\\\"build a Slack app\\\\\\\", \\\\\\\"add a slash command\\\\\\\", \\\\\\\"create a Slack modal\\\\\\\", \\\\\\\"set up Slack Events API\\\\\\\", \\\\\\\"OAuth multi-workspace\\\\\\\", \\\\\\\"submit to Slack Marketplace\\\\\\\", or when slack-app-worker-bee is invoked. Do NOT use ...

Genuine organic social strategy for solo developers, founders, and teams up to ~10 people. No AI-generated posts, no cross-post automation slop, no bought followers. Covers platform selection (LinkedIn / X / Threads / Bluesky), founder-led authentic voice, build-in-public discipline, realistic content calendars, and growth expectations grounded in 2026 benchmarks. Invoke when the user says "help me with social media", "organic growth strategy", "which platform should I post on", "I want to bu...

Public status page specialist: platform selection, incident communication templates, subscriber notifications, compliance. Use when setting up or auditing a status page.

Svelte 5 specialist for runes, snippets, lifecycle, universal reactivity, SvelteKit 2 load/actions/remote functions, and Svelte 4 migration. Use for any .svelte, .svelte.js, or .svelte.ts work.

Tailscale for SvelteKit/Vercel/Neon - tailnets, MagicDNS, ACLs/tags, SSH, subnet routers, exit nodes, private database access, Funnel/Serve, OAuth clients, auth keys, ephemeral CI nodes, key expiry.

Tailwind CSS v4 the framework itself, CSS-first config, @theme, the Vite plugin, v3-to-v4 migration, dark mode, container queries, class ordering. Generic, any codebase, not OSPRY-specific.

TanStack libraries in SvelteKit (Svelte 5) - Query SSR/caching/mutations, Table runes state, Form snippet validation, Virtual; states plainly that Router and Start have no official Svelte support.

Tauri 2 engineering for desktop and mobile: updates, migration, IPC, capabilities, sidecars, AI runtimes, persistence, signing, updater, testing, and distribution. Use for Tauri-specific work.

Writing docs well -- the Diataxis framework (tutorial / how-to / reference / explanation), inverted-pyramid prose structure, scannable headings, code-example discipline, the \\\\\\\"what does the reader already know?\\\\\\\" reader-lens, ghostwriting vs voice consistency, and the docs-as-code review workflow. Distinct from library-worker-bee (which owns docs-site architecture and where a doc lives); this stinger owns the craft of writing. Use when the user says \\\\\\\"review this document\\\...

Telegram Bot specialist: Bot API (up to 10.0 / May 2026 including guest mode and managed bots), grammY v1.x (TypeScript, recommended) and aiogram 3.x (Python), webhook vs long-polling decision with quantitative thresholds, Telegram Mini Apps initData validation (HMAC-SHA256 + Ed25519), Telegram Stars payments (mandatory for digital goods in 2026), inline mode, and MTProto escalation via Telethon/TDLib. Invoke when building a new bot, debugging webhook delivery, wiring Mini Apps, implementing ...

Terminal productivity specialist - Bash/Zsh/Fish configuration, modern CLI tools (ripgrep, fd, fzf, bat, eza, zoxide), shell scripting best practices, dotfile architecture, tmux/Zellij setup, and just/make task automation. Use when the user says \\\\\\\"improve my dotfiles\\\\\\\", \\\\\\\"review this shell script\\\\\\\", \\\\\\\"set up tmux\\\\\\\", \\\\\\\"modern CLI tools\\\\\\\", \\\\\\\"bash best practices\\\\\\\", \\\\\\\"just vs make\\\\\\\", or \\\\\\\"help me with my terminal setup\...

Modern TypeScript/Node practice for this Hive's stack. SvelteKit (Svelte 5) on Vercel as the primary case: tsconfig (bundler resolution, verbatimModuleSyntax), typed load functions/form actions/+server.ts, Drizzle type inference, zod vs valibot at boundaries, Vitest plus Playwright, Biome vs ESLint, pnpm and monorepo choice, Node-on-Vercel policy. Also fully supports the legacy npm library/CLI publishing case (Hivemind stack: strict ESM on Node16 resolution, esbuild bundling, zod v3/v4 MCP sp...

Typography and font-loading specialist for web products: variable fonts, Google Fonts vs Fontsource vs self-host, the FOIT/FOUT/FOFT loading story, font-display semantics, fluid type scales via clamp(), vertical rhythm, and the type-token architecture. Use when the user says \\\\\\\\\\\\\\\"set up fonts\\\\\\\\\\\\\\\", \\\\\\\\\\\\\\\"audit our typography\\\\\\\\\\\\\\\", \\\\\\\\\\\\\\\"fix FOIT/FOUT\\\\\\\\\\\\\\\", \\\\\\\\\\\\\\\"build a type scale\\\\\\\\\\\\\\\", \\\\\\\\\\\\\\\"migrat...

Enforces a product's design system end-to-end — opens the source-of-truth folder first, cites governing sections, specifies pixel-perfect deltas in token-named terms, and applies per-library integration rules for shadcn/ui, Mantine, Lucide-react, and Framer Motion. Use when reviewing UI pull requests, authoring component or screen specs, wrapping an external component library, auditing a screen for token/utility/motion violations, or when the user says "is this on-brief?", "review this screen...

The Svelte 5 + SvelteKit UI enforcement and implementation skill for the OSPRY SvelteKit apps (portal, web, wl). Owns shadcn-svelte 1.x (built on Bits UI v2 + Melt UI), Tailwind v4, the @theme token bridge to the existing PRD-071 design tokens, and the white-label brand contract. Use whenever a PR touches a .svelte file's markup or styling, when adding a shadcn-svelte component, when wiring Tailwind v4 utilities, when bridging an existing CSS custom property into @theme, when verifying an age...

Designs, reviews, and heals vector and embedding storage - schema/column design, index selection (HNSW vs IVFFlat, distance operators), hybrid lexical+vector search, migrations, dataset versioning. Neon plus pgvector plus Drizzle is the primary option for this stack; Deep Lake and Qdrant / managed services remain documented alternatives with a selection matrix. Use when the user says \\\\\\\"design this vector table\\\\\\\", \\\\\\\"which index should this use\\\\\\\", \\\\\\\"pgvector or Dee...