congress-stock-tracker

国会山股神线索追踪工具。追踪美国国会议员最新股票交易动态，识别被多位议员集中买入的重仓标的，结合议员所在委员会和近期立法/监管动态，分析交易背后可能的政策事件催化剂，并检测大额交易、密集交易、跨党派共识等异常信号。当用户询问以下问题时使用本技能：(1) 国会议员最近买了什么股票 (2) 哪些股票被议员重仓买入 (...

Security Scan

The skill's code, instructions, and resources are consistent with its stated purpose of scraping public congressional trade disclosures and performing policy/market correlation analysis; it does not request credentials or install arbitrary third‑party code.

Purpose & Capability

Name/description (tracking congressional trading and doing event‑correlation analysis) match the included files: an HTML/RSC scraping script plus local reference docs for committee–industry mappings and analysis templates. There are no unrelated credential or binary requirements.

Instruction Scope

Runtime instructions limit actions to running scripts, reading bundled reference files, scraping public data sources (Capitol Trades, Quiver, senate disclosures) and performing web searches for event correlation. This is coherent, but the SKILL.md explicitly asks the agent to perform '联网搜索' (network searches) which means the agent may query arbitrary web resources beyond the packaged references — expected for this use case but worth noting.

Install Mechanism

No install spec (instruction-only with a local Python script) — lowest-risk distribution; nothing in the files pulls remote installers or writes unexpected binaries to disk.

Credentials

The skill requires no environment variables, no credentials, and no config paths. Network access to public data sources is required and proportionate to the stated purpose. No requests for unrelated secrets are present.

Persistence & Privilege

always:false and user-invocable: true (default) — the skill does not request permanent/global inclusion or elevated agent privileges and does not modify other skills' configs.

Assessment

This skill appears to be internally consistent: it scrapes publicly available congressional trade disclosure pages and analyzes them against the included committee/analysis references. Before installing, consider: (1) scraping etiquette and terms of service — verify that fetching Capitol Trades / Quiver is permitted and respect rate limits; (2) data limitations — many disclosures report ranges and have up to 45‑day delays, so treat outputs as indicative, not definitive investment advice; (3) web searches invoked at runtime may contact arbitrary external sites — if you need to restrict outbound network access, run the skill in a sandboxed environment first; (4) review the full fetch_trades.py (the provided file appears focused on parsing and analysis) for any unexpected network endpoints or logging of outputs to remote servers; and (5) if you will rely on this for decisions, validate results against primary sources (official disclosures) and consider legal/regulatory/privacy implications of using aggregated political trading data.

🏛️ 国会山股神线索追踪

追踪美国国会议员股票交易，挖掘重仓标的，分析政策线索。

工作流程

用户提问 → 判断查询类型 → 获取交易数据 → 分析处理 → 生成报告

1. 判断查询类型：

   • 全景扫描（"最近议员们买了什么"）→ 获取全量数据 + 重仓分析
   • 定向追踪（"佩洛西最近的交易"）→ 按议员/标的筛选
   • 异常猎手（"有什么异常交易"）→ 异常检测模式
   • 深度分析（"为什么这么多人买 NVDA"）→ 事件关联分析

2. 获取数据：运行 scripts/fetch_trades.py

3. 分析数据：结合 references/ 中的知识进行深度解读

4. 生成报告：按模板输出结构化表格 + 分析解读

数据获取

运行数据抓取脚本获取最新交易记录：

# 基础用法：获取最近30天数据并执行分析
python scripts/fetch_trades.py --days 30 --analyze

# 筛选特定议员
python scripts/fetch_trades.py --politician "Pelosi" --analyze

# 筛选特定股票
python scripts/fetch_trades.py --ticker NVDA --analyze

# 筛选大额交易（>$100K）
python scripts/fetch_trades.py --min-amount 100000 --analyze

# 仅看参议院
python scripts/fetch_trades.py --chamber senate --analyze

# 保存结果到文件
python scripts/fetch_trades.py --analyze --output trades.json
python scripts/fetch_trades.py --output trades.csv

脚本输出 JSON 格式，包含：

• trades：筛选后的交易记录列表
• concentration：重仓标的排行（--analyze 模式）
• anomalies：异常交易信号（--analyze 模式）
• meta：查询元数据

数据源优先级：Capitol Trades → Quiver Quantitative API → 参议院官方披露

分析流程

重仓标的分析

拿到数据后，对被多位议员买入的标的进行深度分析：

1. 读取 references/committee_industry_map.md，查找买入议员所在委员会
2. 判断标的行业是否在委员会管辖范围内（高关联 = 更强信号）
3. 搜索近期相关立法、听证会、监管动态（使用联网搜索）
4. 按 references/analysis_framework.md 中的信号强度评估体系打分
5. 生成事件假设和展望

异常交易检测

脚本自动检测三类异常：

• 🔴 大额交易：单笔 ≥ $500K
• 🟡 密集交易：同一议员短期内对同一标的交易 ≥ 3 次
• 🔴 跨党派共识：两党 ≥ 3 位议员同时买入同一标的

事件关联分析

读取 references/committee_industry_map.md 获取委员会-行业映射，结合以下信息源进行关联：

• 议员委员会任职信息
• 近期国会立法日程
• 监管机构（FDA/FCC/EPA/SEC）近期动态
• 地缘政治事件

详细分析模板和评估框架见 references/analysis_framework.md。

输出要求

每次输出必须包含：

1. 交易总览表格：议员、标的、买/卖、金额、日期、党派
2. 重仓排行：被最多议员买入的标的排名
3. 深度分析：Top 3-5 标的的事件关联解读
4. 异常警报：检测到的异常交易信号
5. ⚖️ 免责声明：必须提醒用户数据存在披露延迟（最长45天），分析仅供参考不构成投资建议

完整报告模板见 references/analysis_framework.md 的"输出报告结构"章节。

注意事项

• 国会议员交易披露存在 最长 45 天延迟，数据并非实时
• Capitol Trades 网站结构可能变化，如抓取失败会自动切换备用数据源
• 交易金额以范围形式披露（如 $1,001-$15,000），非精确数字
• 分析结论是基于公开数据的推测，不代表确定性事件