菜单
Use when converting sports evidence into probability estimates, squad recommendations, risk-adjusted rankings, or market comparisons, especially when the answer should remain explainable and uncertainty-aware in any agent runtime.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Use when reading, updating, or governing long-lived national-team markdown memory files so agent runs keep reusable football context without mixing stale facts, speculative claims, and append-only observations.
Use when building or extending the project's web search, crawling, scraping, parsing, or source-normalization systems for sports, fantasy, and prediction workflows, especially when reliability, provenance, freshness, and site-compliance matter across development and production runtimes.
Use when gathering football, sports, fantasy, or prediction-market intelligence from multiple sources, especially when the work requires source comparison, freshness checks, provenance tracking, and extraction of structured evidence for downstream agents in any runtime.
Use when coordinating multiple AI agents for sports, fantasy, or market-analysis tasks, especially when the work benefits from role-based decomposition, disagreement analysis, critique loops, and a final synthesized recommendation across development or production runtimes.
Use when designing or implementing the web interface for this project, especially when the goal is a trading-terminal aesthetic that still feels welcoming, readable, and simple for everyday users, and when agent outputs must be presented clearly inside the product.
| name | forecasting-desk |
| description | Use when converting sports evidence into probability estimates, squad recommendations, risk-adjusted rankings, or market comparisons, especially when the answer should remain explainable and uncertainty-aware in any agent runtime. |
Use this skill when the task is to turn evidence into a forecast or ranked choice.
{
"option": "",
"probability": 0,
"confidence": 0,
"drivers": [],
"risks": [],
"change_triggers": []
}
{
"squad": [],
"captain": "",
"vice_captain": "",
"constraints_used": [],
"core_reasons": [],
"main_risks": [],
"alternatives": []
}