Builder skill. An element of this node asked for a capability it lacks: your input says «Element <id> needs a capability: «…» — need need-…», or the architect answered yes or no about building one, or an element you led to build it said it is done. Find who already does it by the A2A cards and tell the element; nobody does — ask the architect once whether to build it, and where; on yes lead the owner element to build it as a door, check its card names it, tell the asking element it is ready.
Work skill. The node's checked helpful links — a service of a kind the person is about to need (a VPS to keep the node on, later others), each with what it is for, an honest verdict and its link, partner links disclosed. Load it when a person's goal needs an outside service (a server, hosting, a provider) — together with shape-the-need — when an element's agent asks which service to use, and when the owner says «добавь ссылку», «вот моя реферальная ссылка», «убери ссылку».
Work skill. The person is stuck in someone else's interface — DNS records at a domain registrar, Resend, Google Cloud credentials, a form on a site, a button they cannot find in Google Docs — or asks you to do something there for them («помоги настроить DNS», «получи ключи Google», «не могу найти кнопку»). You do it in their Chrome with the Claude in Chrome tools — on this computer, or on a dedicated server their Chrome on any computer through their Claude account: they sign in, you fill in and set up, irreversible steps stay theirs. Without Chrome and the extension you say what to install, or lead them by words step by step.
Rewrite AI-sounding text so it reads like the writer without changing what it says. Use when editing or reviewing prose for AI tells: not-X-but-Y contrasts, one-line closers, staged openers, forced triads, dashes everywhere, inflated claims, sales language, stock AI words, bold labels, or filler. Based on Wikipedia's "Signs of AI writing."
Both natures. The field of people near the architect — a family, a team, a community — who should get the node's help and talk to it from Telegram without ever opening a computer. The facts the method of shape-the-need weighs when the architect speaks of other people or asks who else the node can serve, and how the work runs once the shape is chosen. Load it together with shape-the-need, never instead of it.
Craft-first interface design for dashboards, admin panels, SaaS apps, tools, settings pages, data interfaces, and interactive products. Use when designing, building, reviewing, auditing, or refining product UI where visual craft, layout hierarchy, tokens, states, visual direction, or design-system consistency matter. Not for marketing pages, landing pages, campaigns, or brand-only work.
Both natures. Build and keep a small, honest picture of the person you work for: first contact, saving, updating, forgetting. Load it at the first contact of a session with no picture in your memory, when the person corrects how you speak or work, and when they say «что ты обо мне знаешь», «забудь» or «не запоминай». Keeping something for later («запомни», «напомни о») — use-deep-memory.
Read or change the deployed app's settings — brand name, description, URL, support email, SEO/OpenGraph, PWA, analytics, structured data — and the app's language set. Use when the owner says things like "change my description to 'Рога и копыта'", "set the site URL", "turn on the Organization schema", or "add French". Writes go through the validated setter (the panel's config API) — never raw JSON. Self-sufficient: reads work from the on-disk file alone; writes need the setter, which is registered in every agent.
Work process. Where the node's money goes and whether it lasts today — answered from the node's ledger of paid calls with no model and without reading a key. Load it when the person asks «за что уходят мои деньги», «сколько я трачу», «кто тратит ключ OpenAI», «хватит ли на сегодня», «какой баланс», «почему перестал работать голос/перевод», when `npm run person` or `npm run where` prints a red money line, or when the person asks for a daily budget of a service.
Build process. The node type — a single project (one owner, every element under the core's CONFIG) or an app factory (hosting of projects of different owners). Load it when the person asks what kind of node this is, speaks about hosting other people's sites, an app factory, projects for clients, one owner or one wallet, or when you create an element and need to know how it will look.
Both natures. The field of a person's own effectiveness — their health, training, food, money, investments, learning and the dates of their life. The facts the method of shape-the-need weighs when a person wants help for themselves, and how the work runs once they chose its shape. Load it together with shape-the-need, never instead of it; for people near the architect — skill integrator.
Make a BUILD-TIME environment variable survive a production redeploy. Use whenever a feature needs a value baked at build — any NEXT_PUBLIC_* (inlined into the browser bundle), the language set, Stripe keys + product ids, a custom API/DB URL, a feature flag, analytics IDs. Write the value into the slot's own app/.env.local through the proper setter, then trigger a rebuild: the slot-scoped build bakes the slot's own .env.local. Use when adding payments, a new integration key, a feature flag, or any value the app reads at build time — and to avoid the silent "saved but the app never sees it" bug. Self-sufficient: works for a lone agent.
Putting a page into the TOP MENU or the FOOTER, and deciding its order. Load it whenever the owner names a page together with a place in the navigation — "add a subscriptions page to the top menu", "put the todo list in the footer", "this should be in the menu", "reorder the footer links". It carries what is easy to get wrong here: whether such a page already exists, whether it may appear in a public menu at all, and which of the two menu sources is actually deciding what the visitor sees right now. The last one fails silently — the page ships, the menu never changes, and nothing reports an error.
Work process. A request may break a provider's rules (many accounts, mass messages, getting around a CAPTCHA or a paywall, scraping a site that forbids it, fake reviews, someone else's brand or identity, prompt injection), a model refused or a call was blocked, the person asks about warnings, limits, bans or appeals, or a batch or scheduled job is about to run on the subscription or the node's agents are to be given to other people. Name the rule, its price and the lawful way — in a few sentences, never a rewording.
Build process. Release a new version of an element template (custom or native) and roll its files out to the node's elements: a working clone in the core's workshop, the template's whole prebuild, a tag on GitHub, the pin in the core, then the same files into every element of that kind with its next patch tag and pin. Load it only when the owner asks to change a template or to bring a template's change into the existing elements; never on your own initiative.
Both natures. The field of research — how an outside product, site, SDK, API or market works, what it solves and how this node can integrate with it, proven by real tests and kept for the next research and the next comparison. The facts the method of shape-the-need weighs when a person asks to research, study, compare or check something outside the node, and how the work runs once the person chose its shape. Load it together with shape-the-need, never instead of it.
Build process. Review a finished page of an element against its approved design: screenshots and the element's page guard without a model first, then a verdict in the owner's three grades with named sections — 1 «replace X → Y», 2 «move X and Z», 3 «looks great» — recorded next to the owner's verdict. Load it when the owner asks to assess, review or check a page of an element, a landing after an element agent's work («look at what the agent built», «grade this page»), and before any verdict on whether a page's sections are right. The rule that changes behaviour: a review by a model costs tokens and runs only after the owner's yes with a price counted for this page; the instrument (screenshots, the guard) is free and runs first.
Both natures. Hand a request to the element of this node that does it — found by its role — with its agent's Telegram bot or terminal; never do its work yourself. Load it when the person asks for something an element does (sign-in, data, settings, design, a site, a custom element's job), asks who does what, or names an element.
Builder skill. The main agent's review of its own work at the architect's request — find in your session where you stumbled and propose at most three changes to your skills or an entry of section 9 of CLAUDE.md, checked against the history of earlier changes and decisions. Load it ONLY when the architect asked for a review («review», «разбор» — in any language) in Telegram or in the terminal. Never without that request.
Manages shadcn components and projects — adding, searching, fixing, debugging, styling, and composing UI, including chat interfaces. Provides project context, component docs, and usage examples. Applies when working with shadcn/ui, component registries, presets, --preset codes, or any project with a components.json file. Also triggers for "shadcn init", "create an app with --preset", or "switch to --preset".
A folder with a SKILL.md file: a name, a description of when to use it, and instructions. Claude loads a skill only when the task matches its description.
How do I use one I find here?
Copy the folder into your project's .claude/skills/ directory, or into your own skills folder to use it everywhere.
What do the warnings mean?
We read each file for commands that read secrets, delete things or pipe downloads into a shell, and say so before you copy it. No warning is not a promise that a file is safe.
Which skills worked for people?
Open a skill to see its discussion. Reports from people and their agents are coming.