convez.ai
SkillBusca & PesquisaIntermediário
jiahongc

Travel Points & Rewards Skills

jiahongc/travel-points-rewards-skills

Briefings de destinos com pesquisa live para agentes de IA

Ver no GitHub
InicialPluginScore · de 0 a 100

Skill que transforma seu agente de IA em assistente de pesquisa de viagens. Oferece briefings detalhados de cidades com pesquisa em tempo real, cobrindo clima, bairros, gastronomia, segurança, transporte e orientações sobre voos e pontos. Disponível em inglês e chinês simplificado.

Instalação

$npx convez-ai jiahongc/travel-points-rewards-skills

1 comando no terminal: detecta o formato do repositório (marketplace ou plugin avulso) e instala via Claude Code. Requer Node 18+ e o Claude Code instalado.

Métricas

Estrelas
2
Forks
0
Acompanhando
0
Saúde
28%
Último commit
há 2 meses
Versão
—
Issues abertas
0
Contribuidores
1

Casos de uso

  • Planejar viagens com informações atualizadas sobre clima, segurança e transporte
  • Pesquisar destinos considerando pontos de viagem e orientações de voos
  • Gerar guias de cidades estruturados e prontos para salvar/compartilhar
  • Obter recomendações culturais e gastronômicas contextualizadas

Veredito

Skill prática e bem-estruturada para quem viaja e usa agentes de IA. Preenche uma lacuna real entre respostas genéricas de chat e pesquisa de viagem útil; a implementação em múltiplas plataformas e linguagens demonstra maturidade.

README

A tradução para o português deste README ainda está na fila.

Travel Points & Rewards Skills

Travel Points & Rewards Skills turns your AI coding agent into a travel research assistant you can query on demand.

It currently includes two destination-briefing skills for Claude Code, Codex, Open Claude, and OpenClaw / ClawHub:

  • /travel-city for English output
  • /travel-city-chinese for Simplified Chinese output with paired Chinese + English place and food names

Both skills cover destination overview, timing, neighborhoods, food, safety, transportation, and optional flight / points context.

Without Travel Points & Rewards Skills

  • You ask for a city guide and get a generic answer based on stale model knowledge
  • Flight and points guidance is vague or missing
  • Seasonal advice is not tied to your actual month of travel
  • Safety and transit details are inconsistent
  • You end up manually turning a chat response into something printable

With Travel Points & Rewards Skills

SkillLanguageWhat it does
/travel-cityEnglishFull city briefing with live research: overview, weather timing, neighborhoods, attractions, events, food, norms, safety, transport, and optional flight / points guidance
/travel-city-chineseSimplified ChineseChinese-first city briefing for Chinese readers, with paired Chinese + English names for places, foods, and transit terms, plus both U.S. and Chinese passport visa guidance by default

Demo

English:

/travel-city Sydney in August from New York

Chinese:

/travel-city-chinese Bergen

Typical output includes:

  • City overview
  • Recent history
  • Best time to visit
  • Top neighborhoods and nearby cities
  • Things to do
  • Popular events
  • Food and dining
  • Cultural norms
  • Safety and security
  • Getting around
  • First-time traveler tips
  • Confidence notes
  • Getting there, when an origin city is provided

Who This Is For

This repo is for travelers, points users, and trip planners who want better destination research than a generic chat answer.

It is not meant to be a casual “top 10 things to do” prompt. The skills are designed as research workflows: they use live search, prioritize primary sources, separate confirmed from unconfirmed information, and produce output that is easier to save, review, and share.


Install

Requirements: Git and any agent that supports skill-style prompts. The repo currently supports Claude Code, Codex, Open Claude, and OpenClaw / ClawHub.

Claude Code

mkdir -p ~/.claude/skills
cp -R .claude/skills/travel-city ~/.claude/skills/travel-city
cp -R .claude/skills/travel-city-chinese ~/.claude/skills/travel-city-chinese

Codex

mkdir -p ~/.codex/skills
cp -R .agents/skills/travel-city ~/.codex/skills/travel-city
cp -R .agents/skills/travel-city-chinese ~/.codex/skills/travel-city-chinese

Open Claude

mkdir -p ~/.agents/skills
cp -R .agents/skills/travel-city ~/.agents/skills/travel-city
cp -R .agents/skills/travel-city-chinese ~/.agents/skills/travel-city-chinese

OpenClaw / ClawHub

If your client is already pointed at this repository, you can use the checked-in skill directories directly:

Repository layout:

.claude/skills/travel-city/
  SKILL.md

.claude/skills/travel-city-chinese/
  SKILL.md

.agents/skills/travel-city/
  SKILL.md

.agents/skills/travel-city-chinese/
  SKILL.md

clawhub/travel-city/
  SKILL.md

clawhub/travel-city-chinese/
  SKILL.md

Use:

  • .claude/skills/... for Claude Code
  • .agents/skills/... for Codex and Open Claude
  • clawhub/... for OpenClaw / ClawHub packaging

Notes:

  • .agents is a tool convention used by Codex-style agent clients, not a universal standard
  • Claude Code does not use agents/openai.yaml in this repo

Setup

For live research, you can optionally set a Brave Search API key:

export BRAVE_API_KEY=your_key_here

Get a key at brave.com/search/api. If you do not set one, OpenClaw-compatible runtimes can use their built-in live web search tools instead.

You need:

  • network access for live web research
  • optional: BRAVE_API_KEY, curl, and gunzip if you want to prefer Brave Search API from the shell

Output Style Differences

/travel-city

  • English output
  • U.S.-passport-first framing unless the user specifies another nationality
  • Standard English travel-guide naming

/travel-city-chinese

  • Simplified Chinese output
  • Default visa guidance for both U.S. and Chinese passports when relevant
  • Chinese-first naming with paired English / local names on first mention, such as 卑尔根(Bergen)
  • Better fit for Chinese travel-planning tone and Chinese-platform source supplementation

Source Policy

Both skills prioritize:

  • Primary: official tourism boards, government travel advisories, immigration / visa pages, CDC travel guidance
  • Secondary: Lonely Planet, The Points Guy, Skyscanner, Google Flights, Trip.com / 携程, 马蜂窝, 穷游, 飞猪, 航旅纵横, Numbeo, Rome2Rio, XE, Grokipedia
  • Tertiary: TripAdvisor, Wikipedia

Both skills explicitly avoid Reddit, X/Twitter, Facebook, Instagram, TikTok, Quora, Medium, and personal blogs.

Grokipedia is intended as a city-guide / reference source for destination context, but not for safety advisories, live pricing, or official transport details.

When platform summaries conflict with official immigration or safety guidance, official sources win.


Roadmap-Friendly Notes

This repo is intentionally laid out so more travel and points skills can be added later without changing the install story.

  • Keep .claude/skills/..., .agents/skills/..., and clawhub/... aligned by capability
  • Keep each skill self-contained in its own directory
  • Treat SKILL.md as the canonical behavior spec
  • Keep examples and install commands in the README in sync with the actual shipped directories

License

MIT

Ver o original no GitHub

Bom para

Relacionados