Help us improve
Share bugs, ideas, or general feedback.
From stock-deep-analyzer
Simulates a 65-investor panel voting on a stock using dimensions.json and raw_data.json, covering 9 investment styles (value, growth, macro, technical, quant, etc.) and outputting Pydantic Signals with consensus statistics.
npx claudepluginhub wbh604/uzi-skill --plugin stock-deep-analyzerHow this skill is triggered — by the user, by Claude, or both
Slash command
/stock-deep-analyzer:investor-panelThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
读取以下输入:
assets/investor-cards.jsonreferences/group-a-classic-value.mdreferences/group-b-growth.mdreferences/group-c-macro-hedge.mdreferences/group-d-technical.mdreferences/group-e-china-value.mdreferences/group-f-china-youzi.mdreferences/group-g-quant.mdreferences/group-i-serenity.mdreferences/quotes-knowledge-base.mdreferences/serenity-voice.mdRoutes investor reasoning overlays (Buffett, Graham, Lynch, etc.) grounded in LLMQuant Data for valuation and fundamental analysis workflows.
Chains InvestSkill modules for structured stock research: business analysis, valuation, market signals, technicals into unified investment thesis with composite score. Use for due diligence.
Stock deep-analysis workflow covering A-shares, HK, US equities. Generates 22-dimension data, 65-expert quantitative review, 6 institutional valuation models (DCF/Comps/LBO/3-Stmt/Merger), and 7 research deliverables with Bloomberg-style HTML reports.
Share bugs, ideas, or general feedback.
读取以下输入:
.cache/{ticker}/dimensions.json — 19 维评分.cache/{ticker}/raw_data.json — 原始数据scripts/lib/investor_db.py — 65 人元数据scripts/lib/seat_db.py — 22 位游资射程规则输出:
.cache/{ticker}/panel.json — 50 个 Signal + 投票统计每个投资者必须返回严格 JSON:
{
"investor_id": "buffett",
"name": "巴菲特",
"group": "A",
"avatar": "avatars/buffett.svg",
"signal": "bullish | neutral | bearish",
"confidence": 87,
"score": 82,
"verdict": "强烈买入 | 买入 | 关注 | 观望 | 等待 | 回避 | 不达标 | 不适合",
"reasoning": "1-3 句具体逻辑",
"comment": "用该投资者语言风格的金句 1-2 句",
"pass": ["..."],
"fail": ["..."],
"ideal_price": 16.20,
"period": "3-5 年"
}
Confidence 校准规则:
from lib.investor_db import INVESTORS, by_group
from lib.seat_db import SEATS, is_in_range
fields 白名单对 22 位游资,先用 is_in_range(nickname, ticker_features) 判断是否在射程内:
signal: "neutral", verdict: "不适合", confidence: 90, comment: "{nick}的射程是{style},这只票不在风格内。"{
"panel_consensus": (bullish_count / 50) * 100,
"vote_distribution": Counter(verdict for i in investors),
"signal_distribution": Counter(signal for i in investors),
"investors": [...]
}
按需读取下列 references:
| 组 | 文件 | 人数 |
|---|---|---|
| A 经典价值 | references/group-a-classic-value.md | 6 |
| B 成长投资 | references/group-b-growth.md | 4 |
| C 宏观对冲 | references/group-c-macro-hedge.md | 5 |
| D 技术趋势 | references/group-d-technical.md | 4 |
| E 中国价投 | references/group-e-china-value.md | 6 |
| F 游资 | references/group-f-china-youzi.md | 22 |
| G 量化系统 | references/group-g-quant.md | 3 |
每次生成 comment 之前必须读 references/quotes-knowledge-base.md 查找该投资者的真实公开原话和"风格"字段。这是知识库 single source of truth。
每位投资者的 comment 字段必须像他本人:
每组 reference 文件末尾有 3-5 句真实公开语录作为 few-shot。