Timerise Skills

The index of Agent Skills published by Timerise. Each skill is a self-contained repository that teaches an agent to build one production-grade module for a Next.js App Router app, on the customer's own stack, the way our engineers build it after shipping it many times. Agent Skills is an open format, so the same folder works in Claude Code, Codex CLI, Gemini CLI and the other skills-compatible agents.

The skills are how we build the standard modules of every custom booking system we deliver. They are written by our own senior engineers from six years of building booking systems: each one is a member of the team's own way of building a module they have shipped many times, written down, with the properties it has to hold stated as what the build and the tests verify. None of it is synthetic: no skill was generated by a model from a prompt. They are published so the code you receive from us is built on parts you can read, and so anyone running a skills-compatible agent can build on the same parts.

Skills

Skill	Version	Builds	Backend
blog-markdown	0.1.5	File-based multilingual blog with localized slugs, tag pages, related posts, RSS, sitemap, hreflang and a CI content validator	Repository files, no CMS
booking-kiosk	0.1.4	Self-service touchscreen kiosk: walk-up booking flow, on-screen keyboard, pay-at-counter or pay-by-QR, server-priced idempotent booking API, LAN failover contract	Backend- and payment-agnostic KioskBackend seam, Firestore reference implementation
digital-signage	0.1.6	In-venue screens: media library, per-screen playlists, device pairing by PIN or URL, fullscreen TV player with stall recovery and health monitoring	Firestore or Supabase
help-center-markdown	0.2.8	Markdown-backed help center: category, tag and article pages, ranked client-side search, locale fallback, JSON-LD, sitemap, CI content validator	Repository files, no CMS
island-mode-server	0.1.4	On-premise fallback server: live RxDB replica of a site's Firestore slice, LAN takeover when the internet drops, idempotent reconnect flush, HMAC hardware auth	Firestore cloud, RxDB on the local box
ksef	1.2.4	KSeF API 2.0 integration for Poland's mandatory e-invoicing: token auth, invoice encryption, interactive and batch sending, UPO receipts, purchase-invoice sync, QR codes	Postgres (Neon/Supabase) on Vercel
stripe-connect-subscriptions	0.1.7	Stripe Connect marketplace settlement and platform subscription billing: split charges, escrow and reserves, account onboarding, off-session billing with dunning, ledger reconciliation	Backend-agnostic store adapter

The version is each skill's latest release as recorded in its CHANGELOG.md; the skill's own repository is the authoritative copy and carries the full feature list. Every skill is versioned on its own line under one release standard: a vX.Y.Z tag, a chore(release): X.Y.Z commit, a Keep a Changelog entry in CHANGELOG.md, and a GitHub Release per tag carrying that entry. More skills are coming shortly. Each is published once the engineer who owns it has audited it against a module they have shipped, and listed here the same day. Watch this repository for additions.

Install

The skills.sh CLI installs a skill into every skills-compatible agent it detects, including Claude Code, Codex CLI and Gemini CLI, with one command:

npx skills add timerise-ai/help-center-markdown

Name the agents instead with -a, for example npx skills add timerise-ai/ksef -a claude-code -a codex.

Manual install

Nothing in a skill is Claude-specific. Each one is a plain Agent Skills folder, SKILL.md plus markdown references with no file that calls a model, so cloning it into an agent's skills directory is all an install is. For Claude Code, one skill:

git clone https://github.com/timerise-ai/help-center-markdown.git ~/.claude/skills/help-center-markdown

All of them:

for skill in \
  blog-markdown \
  booking-kiosk \
  digital-signage \
  help-center-markdown \
  island-mode-server \
  ksef \
  stripe-connect-subscriptions; do
  git clone "https://github.com/timerise-ai/$skill.git" ~/.claude/skills/"$skill"
done

To scope a skill to a single project instead, clone it into that project's .claude/skills/ directory. For another agent, clone into that agent's skills directory, or symlink the copy already cloned for Claude Code so one git pull updates every agent. Update a skill with git pull in its directory.

Activation

A skill activates automatically when a task matches its description, for example "add a knowledge base with search", "pair a TV to a playlist", "make this markdown blog multilingual", "add a walk-up kiosk with counter payment", "keep the site taking bookings when the internet drops", "why is this invoice rejected with a 430", "why is the connected account never funded". It can also be invoked explicitly with its slash command (/blog-markdown, /booking-kiosk, /digital-signage, /help-center-markdown, /island-mode-server, /ksef, /stripe-connect-subscriptions).

Each host matches a task against the description its own way, so invoke a skill explicitly on a first run rather than assuming it fired. The non-negotiables each skill names are where models diverge most, so confirm the ones that matter for your module survived the build.

What every skill looks like

The skills share one layout, so an agent that has used one knows how to read the next:

File	Contents
SKILL.md	Entry point: what the module is, the architecture, the critical facts, the hard rules, and a directory of the references below
references/adaptation.md	The seam contract with the host app, covering auth, tenancy, storage, styling, i18n and the domain rename, plus the module's non-negotiables. Where a skill keeps the seam in SKILL.md or its architecture reference instead, SKILL.md says so
references/*.md	One file per concern: data model, backend specifics, routes, UI, operations, extensions. Loaded on demand; the agent reads only what a task needs
references/provenance.md	The engineering ledger: what the audit of the earlier implementation changed and how the templates verify it, what was kept on purpose, and what is new in the skill. In ksef that record is the changelog
assets/	Runnable files a reference points at rather than inlines: in ksef, the TypeScript examples for auth, crypto, the API client, sending and QR codes; in island-mode-server, the passing vitest suite covering HMAC verification, offline tokens, delta fold-out and failover rescan
CHANGELOG.md	Semantic versioning

Three properties hold across the pack:

• Earned, not invented. A skill is not synthetic. It starts from what a senior engineer on the team already knows how to build: a module they have shipped and reshipped over the years. Where it wraps someone else's API, as ksef wraps the Ministry of Finance's, it starts from that vendor's own specification, corrected against what integrating it taught us. Either way every odd-looking part is there for a reason, and provenance.md or the changelog records which one.
• Portable across backends. The data layer sits behind an explicit seam. Firestore, Supabase/Postgres or a custom store adapter are described side by side; the host app's auth, tenancy, styling and i18n stay the host app's.
• A short list of non-negotiables. Each skill names the handful of rules that travel with the module because the module does not hold without them: hashed per-device tokens, summaries rather than whole corpora sent to the client, webhook idempotency, an invoice's seller NIP bound to the authenticating taxpayer. Everything else is adaptable.

How we use them

We build custom booking systems that customers own outright: one fixed price, full code handoff, zero commission, delivered in weeks. Our delivery process runs from a project brief, through a clickable prototype and quote within 48 hours and a two-to-three-week AI-accelerated build, to full handoff with three years of support. It depends on the agent producing the repetitive parts correctly the first time so engineers spend their time on the business logic that makes a system yours. These skills are how the agent is told what "correctly" means for each module. When an engineer finds a better way to build a module on a project, the lesson goes back into the skill so the next project starts from it.

The stack the skills assume is the one we recommend and hand over: Next.js App Router, Supabase or Firebase, Vercel. See the Toolkit for the booking core these modules attach to: the open-source admin panel, booking pages and services catalog, and the API docs.

Contributing

Issues and pull requests belong in the individual skill repositories. Skills are pure markdown with no build step; code blocks in some skills are type-checked in CI, and a repository's CLAUDE.md, where it has one, carries its editing conventions. Read references/provenance.md before simplifying anything.

To propose a new skill, open an issue here describing the module, the kind of system you have shipped it in, and the properties you have learned to build into it. To build one, follow STANDARD.md: the repository layout, the README shape, the writing rules and the release standard every skill in this index meets.

Author

Built and maintained by Timerise, a small senior team from Wrocław, Poland, building custom booking systems you own.

License

Every skill in this index is released under the MIT License, as is this repository.