How a Programmer Built an AI Agent Skill to Watch the 2026 World Cup

Overview

The haizei-worldcup-2026 project is an AI Agent Skill that fetches real‑time data for the 2026 USA‑Canada‑Mexico World Cup from Baidu Sports and outputs standardized JSON consumable by any AI tool. It is implemented in pure Node.js (≈3500 lines, MIT‑licensed, zero external dependencies) and covers 48 teams, 12 groups, 6 data categories, and all 104 matches.

Installation

Install the skill with a single command:

npx skills add https://github.com/wuchubuzai2018/haizei-worldcup-2026

OpenCode/Claude Code users can use the same command. The repository can also be cloned and run as a command‑line tool:

git clone https://github.com/wuchubuzai2018/haizei-worldcup-2026.git
cd haizei-worldcup-2026
node skills/haizei-worldcup-2026-skill/scripts/worldcup-schedule.js today

Data Categories

Team configuration : complete grouping of 48 teams across 12 groups, including host status and qualification.

Schedule : multi‑dimensional filtering by date, group, team, and event.

Match details : analysis, line‑ups, live event stream, statistics (possession, shots, corners, cards), and betting odds from four bookmakers.

Team details : profiles, rosters, historical results, season‑by‑season statistics.

Player details : ability radar, transfer history, honors, news, and dual‑view schedules (national team & club).

Ranking data : group‑stage points, player rankings across 34 metrics, and official FIFA rankings for 211 teams.

Anti‑scraping Measures

The tool rotates 31 real browser User‑Agent strings, fully mimics request headers, and runs with no external dependencies, allowing it to avoid being blocked by the source site.

Example Queries and Outputs

Natural‑language queries are answered within a second. Examples include:

"What is today’s schedule?" – returns the day’s match list.

"Who are the core players of team X?" – returns the team roster sorted by market value.

"Predict the outcome of Brazil vs Morocco" – returns analysis, historical head‑to‑head, injury reports, and betting odds for both sides.

Specific factual outputs demonstrated in the article:

Opening match (Mexico vs South Africa) featured three red cards—the first time in World Cup history that an opening match produced three reds, with the referee from Guatemala.

Canada’s all‑time World Cup record: 2 wins, 8 losses; the 1‑1 draw on 12 June was its first point in 40 years (since 1986).

USA’s 4‑1 win over Paraguay included the tournament’s first own goal.

China’s cultural presence: a Chinese special guest at the opening ceremony and Chinese‑made merchandise sold in Los Angeles.

Workflow Documentation

The project includes 12 real‑scenario workflow documents that map natural‑language intents to script calls. For example, the query "What matches are today?" triggers the schedule script; "Who is the core player of team X?" triggers a team query followed by sorting by transfer value; "Analyze Brazil vs Morocco" pulls analysis, historical results, and injury information for both teams.

Repository

GitHub repository: https://github.com/wuchubuzai2018/haizei-worldcup-2026 (MIT license).