aria-code 4.1.3__py3-none-any.whl

apps/cli/utils/market_detect.py

@@ -0,0 +1,1578 @@
+# Auto-extracted from aria_cli.py — shared market-detection constants & helpers.
+# Multiple handler modules import from here to avoid circular dependencies.
+
+import re as _re_sym
+
+try:
+    from apps.cli.market_universe import (
+        looks_like_unresolved_market_name as _looks_like_unresolved_market_name,
+        resolve_market_mentions as _resolve_market_mentions,
+        resolve_market_symbol as _resolve_market_symbol,
+    )
+except Exception:
+    def _resolve_market_symbol(_message: str) -> str:
+        return ""
+    def _resolve_market_mentions(_message: str, limit: int = 6):
+        return []
+    def _looks_like_unresolved_market_name(_message: str) -> bool:
+        return False
+
+# Regex to find stock symbols in a message: e.g. "AAPL", "苹果AAPL", "TSLA股票", "BTC"
+_STOCK_PATTERN = _re_sym.compile(
+    r'\b([A-Z]{1,5}(?:\.(?:HK|SH|SZ))?)(?:\s*(?:股票|股价|价格|现在|今天|行情|涨跌))?\b'
+    r'|(?:(?:苹果|特斯拉|英伟达|谷歌|亚马逊|微软|腾讯|阿里|百度|字节)).*?(\b[A-Z]{2,5}\b)',
+    _re_sym.UNICODE
+)
+_CRYPTO_WORDS = {"比特币":"BTC","以太坊":"ETH","狗狗币":"DOGE","索拉纳":"SOL","BTC":"BTC","ETH":"ETH"}
+
+_COMPANY_TO_TICKER = {
+    # ══════════════════════════════════════════════════════════════════════
+    # A股 — 银行
+    # ══════════════════════════════════════════════════════════════════════
+    "工商银行": "601398", "工行": "601398",
+    "农业银行": "601288", "农行": "601288",
+    "中国银行": "601988", "中行": "601988",
+    "建设银行": "601939", "建行": "601939",
+    "交通银行": "601328", "交行": "601328",
+    "招商银行": "600036", "招行": "600036",
+    "兴业银行": "601166", "兴业": "601166",
+    "浦发银行": "600000", "浦发": "600000",
+    "民生银行": "600016", "民生": "600016",
+    "平安银行": "000001",
+    "光大银行": "601818",
+    "华夏银行": "600015",
+    "北京银行": "601169",
+    "南京银行": "601009",
+    "宁波银行": "002142",
+    "成都银行": "601838",
+    "苏州银行": "002966",
+    "江苏银行": "600919",
+    "上海银行": "601229",
+    "杭州银行": "600926",
+    "长沙银行": "601577",
+    "重庆银行": "601963",
+    "贵阳银行": "601997",
+    "郑州银行": "002936",
+    "青岛银行": "002948",
+    "西安银行": "600928",
+    "邮储银行": "601658",
+    # ══════════════════════════════════════════════════════════════════════
+    # A股 — 保险
+    # ══════════════════════════════════════════════════════════════════════
+    "中国平安": "601318", "平安": "601318",
+    "中国人寿": "601628", "人寿": "601628",
+    "中国太保": "601601", "太保": "601601",
+    "新华保险": "601336", "新华": "601336",
+    "中国人保": "601319", "人保": "601319",
+    "中国财险": "02328.HK",
+    # ══════════════════════════════════════════════════════════════════════
+    # A股 — 证券
+    # ══════════════════════════════════════════════════════════════════════
+    "中信证券": "600030", "中信": "600030",
+    "华泰证券": "601688", "华泰": "601688",
+    "国泰君安": "601211", "君安": "601211",
+    "海通证券": "600837", "海通": "600837",
+    "广发证券": "000776", "广发": "000776",
+    "东方证券": "600958",
+    "中国银河": "601881", "银河证券": "601881",
+    "招商证券": "600999",
+    "申万宏源": "000166",
+    "兴业证券": "601377",
+    "东北证券": "000686",
+    "方正证券": "601901",
+    "光大证券": "601788",
+    "国信证券": "002736",
+    "中金公司": "601995",
+    "东方财富": "300059",
+    # ══════════════════════════════════════════════════════════════════════
+    # A股 — 能源 / 石油 / 煤炭 / 电力
+    # ══════════════════════════════════════════════════════════════════════
+    "中国石油": "601857", "中石油": "601857",
+    "中国石化": "600028", "中石化": "600028",
+    "中国海油": "600938", "海油": "600938",
+    "中国神华": "601088", "神华": "601088",
+    "中煤能源": "601898",
+    "兖矿能源": "600188",
+    "陕西煤业": "601225",
+    "山西焦煤": "000983",
+    "淮北矿业": "600985",
+    "平煤股份": "601666",
+    "冀中能源": "000937",
+    "华能国际": "600011",
+    "大唐发电": "601991",
+    "华电国际": "600027",
+    "国电电力": "600795",
+    "三峡能源": "600905",
+    "长江电力": "600900", "长电": "600900",
+    "国投电力": "600886",
+    "华润电力": "0836.HK",
+    # ══════════════════════════════════════════════════════════════════════
+    # A股 — 矿业 / 金属 / 有色（关键补充，修复江西铜业误识别）
+    # ══════════════════════════════════════════════════════════════════════
+    "江西铜业": "600362", "江铜": "600362",          # ← 关键修复
+    "铜陵有色": "000630",
+    "云南铜业": "000878",
+    "中国铝业": "601600", "中铝": "601600",
+    "云铝股份": "000807",
+    "神火股份": "000933",
+    "洛阳钼业": "603993", "洛钼": "603993",
+    "紫金矿业": "601899", "紫金": "601899",
+    "赣锋锂业": "002460", "赣锋": "002460",
+    "天齐锂业": "002466", "天齐": "002466",
+    "藏格矿业": "000408",
+    "西部矿业": "601168",
+    "中国黄金": "600916",
+    "山东黄金": "600547",
+    "招金矿业": "1818.HK",
+    "华友钴业": "603799",
+    "格林美": "002340",
+    "寒锐钴业": "300618",
+    "盛屯矿业": "600711",
+    "厦门钨业": "600549",
+    "中国稀土": "000831",
+    "北方稀土": "600111",
+    "盛和资源": "600392",
+    "金钼股份": "601958",
+    "中国宝武": "600018",  # 宝钢股份
+    "宝钢股份": "600018",
+    "鞍钢股份": "000898",
+    "首钢股份": "000959",
+    "华菱钢铁": "000932",
+    "方大特钢": "600507",
+    "马钢股份": "600808",
+    # ══════════════════════════════════════════════════════════════════════
+    # A股 — 新能源 / 光伏 / 储能
+    # ══════════════════════════════════════════════════════════════════════
+    "宁德时代": "300750", "宁德": "300750", "CATL": "300750",
+    "比亚迪": "002594", "BYD": "002594",
+    "隆基绿能": "601012", "隆基": "601012",
+    "阳光电源": "300274",
+    "通威股份": "600438", "通威": "600438",
+    "晶澳科技": "002459",
+    "天合光能": "688599",
+    "晶科能源": "688223",
+    "固德威": "688390",
+    "亿纬锂能": "300014",
+    "璞泰来": "603659",
+    "恩捷股份": "002812",
+    "星源材质": "300568",
+    "科达利": "002850",
+    "天赐材料": "002709",
+    "新宙邦": "300037",
+    "德方纳米": "300769",
+    "当升科技": "300073",
+    "容百科技": "688005",
+    "振德医疗": "603301",
+    "孚能科技": "688567",
+    "国轩高科": "002074",
+    "欣旺达": "300207",
+    "派能科技": "688063",
+    "中创新航": "3931.HK",
+    "蜂巢能源": "未上市",
+    "海辰储能": "未上市",
+    # ══════════════════════════════════════════════════════════════════════
+    # A股 — 汽车 / 零部件
+    # ══════════════════════════════════════════════════════════════════════
+    "上汽集团": "600104", "上汽": "600104",
+    "广汽集团": "601238", "广汽": "601238",
+    "长城汽车": "601633", "长城": "601633",
+    "长安汽车": "000625", "长安": "000625",
+    "东风汽车": "600006", "东风": "600006",
+    "江淮汽车": "600418",
+    "北汽蓝谷": "600733",
+    "宇通客车": "600066",
+    "中通客车": "000957",
+    "福田汽车": "600166",
+    "潍柴动力": "000338",
+    "汇川技术": "300124", "汇川": "300124",
+    "三花智控": "002050",
+    "拓普集团": "601689",
+    "伯特利": "603596",
+    "华域汽车": "600741",
+    "均胜电子": "600699",
+    "德赛西威": "002920",
+    "星宇股份": "601799",
+    "继峰股份": "603997",
+    "岱美股份": "603730",
+    # ══════════════════════════════════════════════════════════════════════
+    # A股 — 半导体 / 芯片 / 科技
+    # ══════════════════════════════════════════════════════════════════════
+    "中芯国际": "688981", "中芯": "688981",
+    "海光信息": "688041",
+    "寒武纪": "688256",
+    "韦尔股份": "603501",
+    "澜起科技": "688008",
+    "中科曙光": "603019",
+    "北方华创": "002371",
+    "中微公司": "688012",
+    "华润微": "688396",
+    "士兰微": "600460",
+    "斯达半导": "603290",
+    "扬杰科技": "300373",
+    "宏微科技": "688711",
+    "芯原股份": "688521",
+    "晶晨股份": "688099",
+    "卓胜微": "300782",
+    "国芯科技": "688262",
+    "瑞芯微": "603893",
+    "全志科技": "300458",
+    "兆易创新": "603986",
+    "北京君正": "300223",
+    "紫光国微": "002049",
+    "华大九天": "301269",
+    "概伦电子": "688206",
+    "广立微": "301095",
+    "沪硅产业": "688126",
+    "有研新材": "600206",
+    "神工股份": "688233",
+    "中环股份": "002129",
+    "捷捷微电": "300623",
+    "华峰测控": "688200",
+    "长川科技": "300604",
+    "精测电子": "300567",
+    "中科飞测": "688361",
+    "拓荆科技": "688072",
+    "芯源微": "688037",
+    "盛美上海": "688082",
+    "至纯科技": "603690",
+    "华亚智能": "003043",
+    "科大讯飞": "002230", "讯飞": "002230",
+    "海康威视": "002415", "海康": "002415",
+    "大华股份": "002236", "大华": "002236",
+    "工业富联": "601138",
+    "立讯精密": "002475", "立讯": "002475",
+    "歌尔股份": "002241", "歌尔": "002241",
+    "蓝思科技": "300433",
+    "三安光电": "600703",
+    "木林森": "002745",
+    "中国软件": "600536",
+    "浪潮信息": "000977", "浪潮": "000977",
+    "中兴通讯": "000063", "中兴": "000063",
+    "烽火通信": "600498",
+    "中国联通": "600050", "联通": "600050",
+    "中国电信": "601728", "电信": "601728",
+    "中国移动": "600941", "移动": "600941",
+    "京东方": "000725", "BOE": "000725",
+    "TCL科技": "000100", "TCL": "000100",
+    "视源股份": "002841",
+    "闻泰科技": "600745",
+    "鸿海精密": "2317.TW",
+    # ══════════════════════════════════════════════════════════════════════
+    # A股 — 消费 / 食品饮料 / 白酒
+    # ══════════════════════════════════════════════════════════════════════
+    "贵州茅台": "600519", "茅台": "600519",
+    "五粮液": "000858",
+    "泸州老窖": "000568", "老窖": "000568",
+    "洋河股份": "002304", "洋河": "002304",
+    "山西汾酒": "600809", "汾酒": "600809",
+    "古井贡酒": "000596", "古井": "000596",
+    "今世缘": "603369",
+    "酒鬼酒": "000799",
+    "迎驾贡酒": "603198",
+    "舍得酒业": "600702",
+    "岩石股份": "600696",
+    "牧原食品": "002714", "牧原": "002714",
+    "温氏股份": "300498", "温氏": "300498",
+    "新希望": "000876",
+    "天邦食品": "002124",
+    "双汇发展": "000895", "双汇": "000895",
+    "海天味业": "603288", "海天": "603288",
+    "中炬高新": "600872",
+    "恒顺醋业": "600305",
+    "绝味食品": "603517",
+    "安井食品": "603345",
+    "千味央厨": "001215",
+    "桃李面包": "603866",
+    "伊利股份": "600887", "伊利": "600887",
+    "蒙牛乳业": "2319.HK", "蒙牛": "2319.HK",
+    "光明乳业": "600597",
+    "三元股份": "600429",
+    "味知香": "605089",
+    "安琪酵母": "600298",
+    "中粮糖业": "600737",
+    "东鹏饮料": "605499",
+    "农夫山泉": "9633.HK",
+    "百润股份": "002568",
+    "华润啤酒": "0291.HK",
+    "青岛啤酒": "600600", "青啤": "600600",
+    "燕京啤酒": "000729",
+    "重庆啤酒": "600132",
+    "海底捞": "6862.HK",
+    "九毛九": "9922.HK",
+    "百胜中国": "9987.HK",
+    "瑞幸咖啡": "LKNCY",
+    "锅圈食品": "2517.HK",
+    # ══════════════════════════════════════════════════════════════════════
+    # A股 — 医疗 / 医药 / 生物
+    # ══════════════════════════════════════════════════════════════════════
+    "恒瑞医药": "600276", "恒瑞": "600276",
+    "迈瑞医疗": "300760", "迈瑞": "300760",
+    "爱尔眼科": "300015", "爱尔": "300015",
+    "药明康德": "603259", "药明": "603259",
+    "片仔癀": "600436",
+    "云南白药": "000538",
+    "同仁堂": "600085",
+    "华润三九": "000999",
+    "东阿阿胶": "000423",
+    "华东医药": "000963",
+    "复星医药": "600196", "复星": "600196",
+    "上海医药": "601607",
+    "国药股份": "600511",
+    "国药一致": "000028",
+    "康龙化成": "300759",
+    "昭衍新药": "603127",
+    "泰格医药": "300347",
+    "凯莱英": "002821",
+    "九洲药业": "603456",
+    "博腾股份": "300363",
+    "美迪西": "688202",
+    "阳光诺和": "688621",
+    "方达控股": "1521.HK",
+    "普蕊斯": "301257",
+    "迪安诊断": "300244",
+    "金域医学": "603882",
+    "艾德生物": "300685",
+    "贝瑞基因": "000710",
+    "华大基因": "300676",
+    "华大智造": "688114",
+    "诺唯赞": "688105",
+    "奥泰生物": "688606",
+    "乐普医疗": "300003",
+    "心脉医疗": "688016",
+    "微创医疗": "0853.HK",
+    "联影医疗": "688271",
+    "惠泰医疗": "688617",
+    "博迈科": "300404",
+    "振德医疗": "603301",
+    "蓝帆医疗": "002382",
+    "康泰生物": "300601",
+    "智飞生物": "300122",
+    "万泰生物": "603392",
+    "沃森生物": "300142",
+    "华兰生物": "002007",
+    "天坛生物": "600161",
+    "博雅生物": "300294",
+    "信立泰": "002294",
+    "科伦药业": "002422",
+    "扬子江药业": "未上市",
+    # ══════════════════════════════════════════════════════════════════════
+    # A股 — 房地产
+    # ══════════════════════════════════════════════════════════════════════
+    "万科A": "000002", "万科": "000002",
+    "保利发展": "600048", "保利": "600048",
+    "金地集团": "600383", "金地": "600383",
+    "招商蛇口": "001979", "招蛇": "001979",
+    "华发股份": "600325",
+    "绿地控股": "600606", "绿地": "600606",
+    "中南建设": "000961",
+    "荣盛发展": "002146",
+    "新城控股": "601155",
+    "滨江集团": "002244",
+    "龙湖集团": "0960.HK", "龙湖": "0960.HK",
+    "碧桂园": "2007.HK",
+    "华润置地": "1109.HK",
+    "恒大集团": "3333.HK", "恒大": "3333.HK",
+    "旭辉控股": "0884.HK",
+    "融创中国": "1918.HK",
+    "美的置业": "3990.HK",
+    # ══════════════════════════════════════════════════════════════════════
+    # A股 — 化工 / 材料
+    # ══════════════════════════════════════════════════════════════════════
+    "万华化学": "600309", "万华": "600309",
+    "华鲁恒升": "600426",
+    "鲁西化工": "000830",
+    "扬农化工": "600486",
+    "沿海能源": "001213",
+    "恒逸石化": "000703",
+    "荣盛石化": "002493",
+    "东方盛虹": "000301",
+    "桐昆股份": "601233",
+    "新凤鸣": "603225",
+    "恒力石化": "600346",
+    "卫星化学": "002648",
+    "龙佰集团": "002601",
+    "中国巨石": "600176",
+    "中材科技": "002080",
+    "科顺股份": "300737",
+    "东方雨虹": "002271", "雨虹": "002271",
+    "三棵树": "603737",
+    "回天新材": "300041",
+    # ══════════════════════════════════════════════════════════════════════
+    # A股 — 机械 / 工业
+    # ══════════════════════════════════════════════════════════════════════
+    "三一重工": "600031", "三一": "600031",
+    "中联重科": "000157", "中联": "000157",
+    "徐工机械": "000425", "徐工": "000425",
+    "柳工": "000528",
+    "上海机电": "600835",
+    "海天精工": "601882",
+    "科德数控": "688305",
+    "创世纪": "300083",
+    "华中数控": "300161",
+    "秦川机床": "000837",
+    "欣旺达": "300207",
+    "宏发股份": "600885",
+    "正泰电器": "601877",
+    "良信股份": "002706",
+    "天正电气": "605066",
+    "思源电气": "002028",
+    "许继电气": "000400",
+    "平高电气": "600312",
+    # ══════════════════════════════════════════════════════════════════════
+    # A股 — 零售 / 商业
+    # ══════════════════════════════════════════════════════════════════════
+    "中国中免": "601888", "中免": "601888",
+    "王府井": "600859",
+    "百联股份": "600827",
+    "永辉超市": "601933",
+    "家家悦": "603708",
+    "步步高": "002251",
+    "名创优品": "MNSO",
+    "泡泡玛特": "9992.HK",
+    # ══════════════════════════════════════════════════════════════════════
+    # A股 — 物流 / 快递
+    # ══════════════════════════════════════════════════════════════════════
+    "顺丰控股": "002352", "顺丰": "002352",
+    "圆通速递": "600233", "圆通": "600233",
+    "申通快递": "002468", "申通": "002468",
+    "韵达股份": "002120", "韵达": "002120",
+    "中通快递": "ZTO",
+    "中远海控": "601919",
+    "招商轮船": "601872",
+    "中国外运": "601598",
+    "白云机场": "600004",
+    "上海机场": "600009",
+    "首都机场": "0694.HK",
+    "深圳机场": "000089",
+    "厦门空港": "600897",
+    # ══════════════════════════════════════════════════════════════════════
+    # A股 — 传媒 / 游戏 / 教育
+    # ══════════════════════════════════════════════════════════════════════
+    "三七互娱": "002555",
+    "完美世界": "002624",
+    "游族网络": "002174",
+    "吉比特": "603444",
+    "恺英网络": "002517",
+    "中青宝": "300052",
+    "芒果超媒": "300413", "芒果": "300413",
+    "华策影视": "300133",
+    "光线传媒": "300251",
+    "中文传媒": "600373",
+    "分众传媒": "002027",
+    "中公教育": "002607",
+    # ══════════════════════════════════════════════════════════════════════
+    # 港股 — 主要蓝筹
+    # ══════════════════════════════════════════════════════════════════════
+    "腾讯控股": "0700.HK", "腾讯": "0700.HK",
+    "阿里巴巴港股": "9988.HK", "阿里港股": "9988.HK",
+    "小米集团": "1810.HK", "小米": "1810.HK",
+    "美团": "3690.HK", "美团点评": "3690.HK",
+    "快手": "1024.HK",
+    "百度港股": "9888.HK",
+    "京东港股": "9618.HK",
+    "网易港股": "9999.HK",
+    "哔哩哔哩港股": "9626.HK", "B站港股": "9626.HK",
+    "携程港股": "9961.HK",
+    "汇丰控股": "0005.HK", "汇丰": "0005.HK", "HSBC港股": "0005.HK",
+    "友邦保险": "1299.HK", "友邦": "1299.HK",
+    "中国人寿港股": "2628.HK",
+    "中国平安港股": "2318.HK",
+    "中国太平": "0966.HK",
+    "太古地产": "1972.HK",
+    "新鸿基地产": "0016.HK", "新鸿基": "0016.HK",
+    "长和": "0001.HK",
+    "长实集团": "1113.HK",
+    "恒基地产": "0012.HK",
+    "九龙仓集团": "0004.HK",
+    "李宁": "2331.HK",
+    "安踏体育": "2020.HK", "安踏": "2020.HK",
+    "申洲国际": "2313.HK",
+    "华润啤酒港股": "0291.HK",
+    "中国移动港股": "0941.HK",
+    "中国联通港股": "0762.HK",
+    "中国电信港股": "0728.HK",
+    "中国铁塔": "0788.HK",
+    "中国建筑": "3311.HK",
+    "中国中铁": "0390.HK",
+    "中国交建": "1800.HK",
+    "吉利汽车": "0175.HK", "吉利": "0175.HK",
+    "比亚迪港股": "1211.HK",
+    "长城汽车港股": "2333.HK",
+    "广汽集团港股": "2238.HK",
+    "东风集团": "0489.HK",
+    "药明生物": "2269.HK",
+    "信达生物": "1801.HK",
+    "百济神州港股": "6160.HK",
+    "君实生物": "1877.HK",
+    "荣昌生物": "9995.HK",
+    "康方生物": "9926.HK",
+    "绿叶制药": "2186.HK",
+    "中生制药": "1177.HK",
+    "石药集团": "1093.HK",
+    "联想集团": "0992.HK", "联想": "0992.HK",
+    "瑞声科技": "2018.HK",
+    "舜宇光学": "2382.HK",
+    "中芯国际港股": "0981.HK",
+    "滔搏": "6110.HK",
+    "泡泡玛特港股": "9992.HK",
+    "名创优品港股": "9896.HK",
+    "周大福": "1929.HK",
+    "六福集团": "0590.HK",
+    "周生生": "0116.HK",
+    "莎莎国际": "0178.HK",
+    "海丰国际": "1308.HK",
+    # ══════════════════════════════════════════════════════════════════════
+    # 中概股（美股上市）
+    # ══════════════════════════════════════════════════════════════════════
+    "阿里巴巴": "BABA", "阿里": "BABA",
+    "百度": "BIDU",
+    "京东": "JD",
+    "拼多多": "PDD",
+    "网易": "NTES",
+    "哔哩哔哩": "BILI", "B站": "BILI",
+    "携程": "TCOM",
+    "微博": "WB", "新浪微博": "WB",
+    "虎牙直播": "HUYA",
+    "斗鱼": "DOYU",
+    "爱奇艺": "IQ",
+    "好未来": "TAL",
+    "新东方": "EDU",
+    "高途": "GOTU",
+    "中通快递美股": "ZTO",
+    "唯品会": "VIPS",
+    "满帮": "YMM",
+    "贝壳": "BEKE",
+    "富途控股": "FUTU",
+    "老虎证券": "TIGR",
+    "小鹏汽车": "XPEV",
+    "蔚来汽车": "NIO", "蔚来": "NIO",
+    "理想汽车": "LI", "理想": "LI",
+    "极氪": "ZK",
+    "零跑汽车": "9863.HK",
+    "滴滴": "DIDIY",
+    "字节跳动": "PRIVATE:ByteDance",
+    "瑞幸咖啡美股": "LKNCY",
+    # ══════════════════════════════════════════════════════════════════════
+    # 美股 — 科技 / 互联网
+    # ══════════════════════════════════════════════════════════════════════
+    "Apple": "AAPL", "苹果": "AAPL", "苹果公司": "AAPL",
+    "Microsoft": "MSFT", "微软": "MSFT",
+    "Alphabet": "GOOGL", "Google": "GOOGL", "谷歌": "GOOGL",
+    "Amazon": "AMZN", "亚马逊": "AMZN",
+    "Meta": "META", "Facebook": "META", "脸书": "META",
+    "Netflix": "NFLX", "奈飞": "NFLX", "网飞": "NFLX",
+    "Tesla": "TSLA", "特斯拉": "TSLA", "特斯拉公司": "TSLA",
+    "NVIDIA": "NVDA", "英伟达": "NVDA",
+    "AMD": "AMD",
+    "Intel": "INTC", "英特尔": "INTC",
+    "Qualcomm": "QCOM", "高通": "QCOM",
+    "Broadcom": "AVGO", "博通": "AVGO",
+    "Texas Instruments": "TXN", "德州仪器": "TXN",
+    "Micron": "MU", "美光": "MU",
+    "Western Digital": "WDC", "西部数据": "WDC",
+    "Seagate": "STX", "希捷": "STX",
+    "Cisco": "CSCO", "思科": "CSCO",
+    "Oracle": "ORCL", "甲骨文": "ORCL",
+    "Salesforce": "CRM",
+    "Adobe": "ADBE", "奥多比": "ADBE",
+    "ServiceNow": "NOW",
+    "Workday": "WDAY",
+    "Snowflake": "SNOW",
+    "Databricks": "未上市",
+    "Palantir": "PLTR", "帕兰提尔": "PLTR",
+    "Cloudflare": "NET",
+    "CrowdStrike": "CRWD",
+    "Palo Alto": "PANW",
+    "Fortinet": "FTNT",
+    "Zscaler": "ZS",
+    "Okta": "OKTA",
+    "Datadog": "DDOG",
+    "MongoDB": "MDB",
+    "Twilio": "TWLO",
+    "Shopify": "SHOP",
+    "Uber": "UBER", "优步": "UBER",
+    "Lyft": "LYFT",
+    "Airbnb": "ABNB", "爱彼迎": "ABNB",
+    "DoorDash": "DASH",
+    "Instacart": "CART",
+    "Spotify": "SPOT",
+    "Pinterest": "PINS",
+    "Snap": "SNAP", "Snapchat": "SNAP",
+    "Twitter": "X", "推特": "X",
+    "Reddit": "RDDT",
+    "Roblox": "RBLX",
+    "Unity": "U",
+    "Zoom": "ZM",
+    "Slack": "未独立",
+    "OpenAI": "PRIVATE:OpenAI",
+    "Anthropic": "PRIVATE:Anthropic",
+    "xAI": "PRIVATE:xAI",
+    "SpaceX": "SPCX", "太空探索技术": "SPCX", "太空探索": "SPCX", "星链": "SPCX",
+    "Starlink": "SPCX", "星链卫星": "SPCX",
+    "ByteDance": "PRIVATE:ByteDance",
+    "Stripe": "PRIVATE:Stripe",
+    "Databricks": "PRIVATE:Databricks",
+    "Canva": "PRIVATE:Canva",
+    "Waymo": "PRIVATE:Waymo",
+    "Klarna": "PRIVATE:Klarna",
+    "Revolut": "PRIVATE:Revolut",
+    "Shein": "PRIVATE:Shein", "希音": "PRIVATE:Shein",
+    "SHEIN": "PRIVATE:Shein",
+    "华为": "PRIVATE:Huawei", "Huawei": "PRIVATE:Huawei",
+    "大疆": "PRIVATE:DJI", "DJI": "PRIVATE:DJI",
+    "蚂蚁集团": "PRIVATE:AntGroup", "支付宝": "PRIVATE:AntGroup",
+    "京东物流": "PRIVATE:JDL",
+    "Arm": "ARM",
+    "TSMC": "TSM", "台积电": "TSM",
+    "Samsung": "005930.KS",
+    # ══════════════════════════════════════════════════════════════════════
+    # 美股 — 金融
+    # ══════════════════════════════════════════════════════════════════════
+    "JPMorgan": "JPM", "摩根大通": "JPM",
+    "Goldman Sachs": "GS", "高盛": "GS",
+    "Morgan Stanley": "MS", "摩根士丹利": "MS",
+    "Bank of America": "BAC", "美国银行": "BAC",
+    "Citigroup": "C", "花旗": "C",
+    "Wells Fargo": "WFC", "富国银行": "WFC",
+    "Berkshire": "BRK-B", "伯克希尔": "BRK-B", "巴菲特": "BRK-B",
+    "BlackRock": "BLK", "贝莱德": "BLK",
+    "Visa": "V",
+    "Mastercard": "MA",
+    "American Express": "AXP", "美国运通": "AXP",
+    "PayPal": "PYPL",
+    "Block": "SQ", "Square": "SQ",
+    "Coinbase": "COIN",
+    "Robinhood": "HOOD",
+    "Charles Schwab": "SCHW",
+    "Interactive Brokers": "IBKR",
+    # ══════════════════════════════════════════════════════════════════════
+    # 美股 — 消费 / 零售
+    # ══════════════════════════════════════════════════════════════════════
+    "Walmart": "WMT", "沃尔玛": "WMT",
+    "Costco": "COST", "好市多": "COST",
+    "Target": "TGT",
+    "Home Depot": "HD", "家得宝": "HD",
+    "Lowe's": "LOW",
+    "Nike": "NKE", "耐克": "NKE",
+    "Adidas": "ADDYY", "阿迪达斯": "ADDYY",
+    "Lululemon": "LULU",
+    "Tapestry": "TPR",
+    "Coca-Cola": "KO", "可口可乐": "KO",
+    "PepsiCo": "PEP", "百事可乐": "PEP",
+    "McDonald's": "MCD", "麦当劳": "MCD",
+    "Starbucks": "SBUX", "星巴克": "SBUX",
+    "Yum Brands": "YUM",
+    "Chipotle": "CMG",
+    "Procter & Gamble": "PG", "宝洁": "PG",
+    "Unilever": "UL", "联合利华": "UL",
+    "Colgate": "CL",
+    "Estee Lauder": "EL", "雅诗兰黛": "EL",
+    "L'Oreal": "LRLCY", "欧莱雅": "LRLCY",
+    # ══════════════════════════════════════════════════════════════════════
+    # 美股 — 医疗 / 制药
+    # ══════════════════════════════════════════════════════════════════════
+    "Johnson & Johnson": "JNJ", "强生": "JNJ",
+    "Pfizer": "PFE", "辉瑞": "PFE",
+    "Merck": "MRK", "默沙东": "MRK",
+    "Eli Lilly": "LLY", "礼来": "LLY",
+    "AbbVie": "ABBV",
+    "Bristol-Myers": "BMY",
+    "Amgen": "AMGN",
+    "Gilead": "GILD",
+    "Biogen": "BIIB",
+    "Moderna": "MRNA", "莫德纳": "MRNA",
+    "BioNTech": "BNTX", "拜恩泰科": "BNTX",
+    "Regeneron": "REGN",
+    "Vertex": "VRTX",
+    "Illumina": "ILMN",
+    "Thermo Fisher": "TMO", "赛默飞": "TMO",
+    "Abbott": "ABT",
+    "Medtronic": "MDT",
+    "Boston Scientific": "BSX",
+    "UnitedHealth": "UNH",
+    "CVS Health": "CVS",
+    # ══════════════════════════════════════════════════════════════════════
+    # 美股 — 能源
+    # ══════════════════════════════════════════════════════════════════════
+    "ExxonMobil": "XOM", "埃克森美孚": "XOM", "埃克森": "XOM",
+    "Chevron": "CVX", "雪佛龙": "CVX",
+    "ConocoPhillips": "COP",
+    "Shell": "SHEL", "壳牌": "SHEL",
+    "BP": "BP", "英国石油": "BP",
+    "TotalEnergies": "TTE", "道达尔": "TTE",
+    "Schlumberger": "SLB",
+    "Halliburton": "HAL",
+    # ══════════════════════════════════════════════════════════════════════
+    # 美股 — 工业 / 航空 / 国防
+    # ══════════════════════════════════════════════════════════════════════
+    "Boeing": "BA", "波音": "BA",
+    "Lockheed": "LMT", "洛克希德": "LMT",
+    "Raytheon": "RTX",
+    "Northrop": "NOC",
+    "General Dynamics": "GD",
+    "Caterpillar": "CAT",
+    "Deere": "DE", "约翰迪尔": "DE",
+    "Honeywell": "HON", "霍尼韦尔": "HON",
+    "3M": "MMM",
+    "GE": "GE", "通用电气": "GE",
+    "Siemens": "SIEGY", "西门子美股": "SIEGY",
+    "ABB": "ABB",
+    "Emerson": "EMR",
+    "Parker Hannifin": "PH",
+    "United Airlines": "UAL", "美联航": "UAL",
+    "Delta": "DAL", "达美航空": "DAL",
+    "American Airlines": "AAL", "美国航空": "AAL",
+    "Southwest": "LUV", "西南航空": "LUV",
+    "FedEx": "FDX", "联邦快递": "FDX",
+    "UPS": "UPS",
+    # ══════════════════════════════════════════════════════════════════════
+    # 欧洲股票
+    # ══════════════════════════════════════════════════════════════════════
+    "LVMH": "MC.PA", "路易威登": "MC.PA", "路易斯威登": "MC.PA",
+    "爱马仕": "RMS.PA",
+    "开云集团": "KER.PA", "古驰": "KER.PA",
+    "香奈儿": "CHNLF",
+    "标致雪铁龙": "STLAM.MI", "Stellantis": "STLAM.MI",
+    "奔驰": "MBG.DE", "梅赛德斯": "MBG.DE", "梅赛德斯奔驰": "MBG.DE",
+    "宝马": "BMW.DE",
+    "大众": "VOW3.DE", "大众汽车": "VOW3.DE", "Volkswagen": "VOW3.DE",
+    "保时捷": "P911.DE",
+    "奥迪": "NSU.DE",
+    "西门子": "SIE.DE",
+    "SAP": "SAP",
+    "ASML": "ASML", "阿斯麦": "ASML",
+    "英飞凌": "IFNNY",
+    "意法半导体": "STM",
+    "诺基亚": "NOK",
+    "爱立信": "ERIC",
+    "汇丰": "HSBC", "汇丰银行": "HSBC",
+    "巴克莱": "BCS",
+    "德意志银行": "DB", "德银": "DB",
+    "法国巴黎银行": "BNPQY", "法巴": "BNPQY",
+    "瑞银": "UBS",
+    "安联": "ALIZY",
+    "安盛": "AXAHY",
+    "飞利浦": "PHG",
+    "雀巢": "NSRGY", "Nestle": "NSRGY",
+    "诺华": "NVS",
+    "罗氏": "RHHBY",
+    "阿斯利康": "AZN",
+    "GSK": "GSK", "葛兰素史克": "GSK",
+    "优尼利华": "UL",
+    "帝亚吉欧": "DEO",
+    "AB InBev": "BUD", "百威": "BUD",
+    "空客": "EADSY",
+    "赛诺菲": "SNY",
+    "道达尔能源": "TTE",
+    # ══════════════════════════════════════════════════════════════════════
+    # 日本股票
+    # ══════════════════════════════════════════════════════════════════════
+    "丰田": "TM", "丰田汽车": "TM", "Toyota": "TM",
+    "本田": "HMC", "本田汽车": "HMC", "Honda": "HMC",
+    "日产": "NSANY", "Nissan": "NSANY",
+    "马自达": "MZDAY",
+    "斯巴鲁": "FUJHY",
+    "三菱": "MMTOF",
+    "索尼": "SONY", "索尼集团": "SONY", "Sony": "SONY",
+    "松下": "PCRFY", "Panasonic": "PCRFY",
+    "日立": "HTHIY", "Hitachi": "HTHIY",
+    "东芝": "TOSYY", "Toshiba": "TOSYY",
+    "富士通": "FJTSY",
+    "NEC": "NIPNF",
+    "京瓷": "KYCCF",
+    "村田制作所": "MURAY",
+    "TDK": "TTDKY",
+    "日东电工": "NTTDF",
+    "信越化学": "SHECY",
+    "JSR": "JSRCF",
+    "东京电子": "TOELY",
+    "佳能": "CAJ", "Canon": "CAJ",
+    "尼康": "NINOY",
+    "软银": "SFTBY", "SoftBank": "SFTBY",
+    "任天堂": "NTDOY", "Nintendo": "NTDOY",
+    "乐天": "RKUNY",
+    "迅销": "FRCOF", "优衣库": "FRCOF",
+    "资生堂": "SSDOY",
+    "花王": "KAOOY",
+    "武田制药": "TAK",
+    "安斯泰来": "ALPMY",
+    "第一三共": "DSNKY",
+    "大塚制药": "OTSKY",
+    "瑞穗金融": "MFG",
+    "三菱日联": "MUFG",
+    "三井住友": "SMFG",
+    # ══════════════════════════════════════════════════════════════════════
+    # 韩国股票
+    # ══════════════════════════════════════════════════════════════════════
+    "三星电子": "005930.KS", "三星": "005930.KS",
+    "SK海力士": "000660.KS", "海力士": "000660.KS",
+    "LG电子": "066570.KS",
+    "LG化学": "051910.KS",
+    "LG能源": "373220.KS",
+    "现代汽车": "005380.KS", "现代": "005380.KS",
+    "起亚": "000270.KS",
+    "现代摩比斯": "012330.KS",
+    "POSCO": "005490.KS", "浦项制铁": "005490.KS",
+    "SK电讯": "017670.KS",
+    "KT": "030200.KS",
+    "Kakao": "035720.KS",
+    "Naver": "035420.KS",
+    "Celltrion": "068270.KS",
+    # ══════════════════════════════════════════════════════════════════════
+    # 台湾股票
+    # ══════════════════════════════════════════════════════════════════════
+    "台积电美股": "TSM", "台积": "TSM",
+    "台积电台股": "2330.TW",
+    "联发科": "2454.TW", "联发科技": "2454.TW",
+    "鸿海": "2317.TW", "鸿海精密": "2317.TW",
+    "台达电": "2308.TW",
+    "日月光": "3711.TW",
+    "联华电子": "2303.TW", "联电": "2303.TW",
+    "华硕": "2357.TW",
+    "宏碁": "2353.TW",
+    "中华电信": "2412.TW",
+    # ══════════════════════════════════════════════════════════════════════
+    # 指数 / ETF
+    # ══════════════════════════════════════════════════════════════════════
+    "纳斯达克100": "QQQ", "纳斯达克": "QQQ", "纳指": "QQQ",
+    "标普500": "SPY", "标普": "SPY", "S&P500": "SPY", "S&P 500": "SPY",
+    "道琼斯": "DIA", "道指": "DIA", "道琼": "DIA",
+    "罗素2000": "IWM",
+    "标普科技": "XLK",
+    "半导体ETF": "SOXX",
+    "沪深300": "000300.SS", "沪深": "000300.SS",
+    "上证指数": "000001.SS", "上证": "000001.SS",
+    "深证成指": "399001.SZ", "深证": "399001.SZ",
+    "创业板指": "399006.SZ", "创业板": "399006.SZ",
+    "科创50": "000688.SS", "科创板": "000688.SS",
+    "中证500": "000905.SS",
+    "中证1000": "000852.SS",
+    "恒生指数": "^HSI", "恒生": "^HSI", "恒指": "^HSI",
+    "恒生科技": "^HSTECH",
+    "国企指数": "^HSCE", "国企": "^HSCE",
+    "日经225": "^N225", "日经": "^N225",
+    "韩国综合": "^KS11", "KOSPI": "^KS11",
+    "台湾加权": "^TWII",
+    "富时100": "^FTSE", "英国富时": "^FTSE",
+    "德国DAX": "^GDAXI", "DAX": "^GDAXI",
+    "法国CAC": "^FCHI", "CAC40": "^FCHI",
+    "欧洲斯托克50": "^STOXX50E",
+    "澳大利亚": "^AXJO",
+    # ══════════════════════════════════════════════════════════════════════
+    # 加密货币
+    # ══════════════════════════════════════════════════════════════════════
+    "比特币": "BTC-USD", "Bitcoin": "BTC-USD",
+    "以太坊": "ETH-USD", "Ethereum": "ETH-USD",
+    "狗狗币": "DOGE-USD", "Dogecoin": "DOGE-USD",
+    "索拉纳": "SOL-USD", "Solana": "SOL-USD",
+    "瑞波币": "XRP-USD", "XRP": "XRP-USD",
+    "Cardano": "ADA-USD", "艾达币": "ADA-USD",
+    "Avalanche": "AVAX-USD",
+    "Polygon": "MATIC-USD",
+    "Chainlink": "LINK-USD",
+    "Uniswap": "UNI-USD",
+    "莱特币": "LTC-USD", "Litecoin": "LTC-USD",
+    "币安币": "BNB-USD", "BNB": "BNB-USD",
+    "USDT": "USDT-USD",
+    "USDC": "USDC-USD",
+    # ══════════════════════════════════════════════════════════════════════
+    # 大宗商品期货
+    # ══════════════════════════════════════════════════════════════════════
+    "黄金": "GC=F", "Gold": "GC=F",
+    "白银": "SI=F", "Silver": "SI=F",
+    "铂金": "PL=F",
+    "钯金": "PA=F",
+    "WTI原油": "CL=F", "原油": "CL=F", "WTI": "CL=F",
+    "布伦特原油": "BZ=F", "布油": "BZ=F",
+    "天然气": "NG=F",
+    "铜期货": "HG=F", "COMEX铜": "HG=F",   # 注意：是期货不是公司
+    "玉米": "ZC=F",
+    "大豆": "ZS=F",
+    "小麦": "ZW=F",
+    "棉花": "CT=F",
+    "咖啡": "KC=F",
+    "可可": "CC=F",
+    "糖": "SB=F",
+}
+
+# ══════════════════════════════════════════════════════════════════════════════
+# Private company profiles — updated as of 2025-Q2
+# Fields: valuation_usd (latest known, $B), rev_est (annual revenue est, $B),
+#   rev_growth (YoY est %), founded, hq, employees (k),
+#   segments (list), comparables (public tickers for peer analysis),
+#   ipo_status, highlights (list of key facts)
+# ══════════════════════════════════════════════════════════════════════════════
+_PRIVATE_COMPANY_PROFILES: dict[str, dict] = {
+    "OpenAI": {
+        "name": "OpenAI",
+        "valuation_usd": 157,
+        "rev_est": 3.7,
+        "rev_growth": 200,
+        "founded": 2015,
+        "founder": "Sam Altman, Greg Brockman, Elon Musk (已离开)",
+        "hq": "San Francisco, California, USA",
+        "employees": 3.5,
+        "segments": ["ChatGPT 订阅 (Plus/Pro/Team/Enterprise)", "API 服务 (GPT-4o, o3等)", "Sora 视频生成", "企业定制部署"],
+        "comparables": ["MSFT", "GOOGL", "META", "AMZN"],
+        "ipo_status": "2025年完成公司结构从非营利转为营利，IPO 预计 2026-2027",
+        "last_funding": "2025-Q1 $40B 融资，SoftBank 领投，估值 $157B",
+        "highlights": [
+            "ChatGPT 月活超 5 亿，企业客户超 100 万",
+            "2024 年 ARR 超 $3.7B，增速 >200% YoY",
+            "与微软 $13B 战略合作，Azure OpenAI 是主要分发渠道",
+            "o3/o4 系列推理模型引领 AI 能力竞争",
+        ],
+        "risks": [
+            "持续巨额亏损（计算成本极高，2024 运营亏损 ~$5B）",
+            "Google Gemini、Anthropic Claude、Meta LLaMA 激烈竞争",
+            "人才流失风险（联创及核心研究员多次离职）",
+            "监管压力：全球 AI 监管法规趋严",
+        ],
+    },
+    "ByteDance": {
+        "name": "ByteDance (字节跳动)",
+        "valuation_usd": 300,
+        "rev_est": 120,
+        "rev_growth": 30,
+        "founded": 2012,
+        "founder": "张一鸣 (Zhang Yiming)",
+        "hq": "北京，中国 / 新加坡（国际总部）",
+        "employees": 110,
+        "segments": ["抖音/TikTok 广告", "今日头条/西瓜视频", "飞书企业办公", "游戏/教育/电商"],
+        "comparables": ["META", "GOOGL", "SNAP", "PINS"],
+        "ipo_status": "屡次推迟；TikTok 美国封禁风险悬而未决；整体 IPO 无明确时间表",
+        "last_funding": "二级市场估值 ~$300B；上一轮融资 2021 年 $5B",
+        "highlights": [
+            "TikTok 月活超 17 亿，广告收入超 $200 亿",
+            "抖音国内广告收入 2024 超 $500 亿人民币",
+            "飞书企业服务增速快，但仍处亏损扩张期",
+            "整体集团收入约 $1200 亿（2024 est.），接近 Meta",
+        ],
+        "risks": [
+            "TikTok 美国强制出售/封禁立法风险",
+            "中国数据安全监管，出海受限",
+            "广告市场周期性，高度依赖广告营收",
+        ],
+    },
+    "Anthropic": {
+        "name": "Anthropic",
+        "valuation_usd": 61,
+        "rev_est": 1.5,
+        "rev_growth": 300,
+        "founded": 2021,
+        "founder": "Dario Amodei, Daniela Amodei (原 OpenAI)",
+        "hq": "San Francisco, California, USA",
+        "employees": 1.5,
+        "segments": ["Claude API (Haiku/Sonnet/Opus)", "Claude.ai 订阅", "企业/政府部署"],
+        "comparables": ["MSFT", "GOOGL", "META"],
+        "ipo_status": "无公开 IPO 计划，处于高速成长期",
+        "last_funding": "2025 Google $3B + Amazon $4B 战略投资，估值 $61B",
+        "highlights": [
+            "Claude 3.5 系列被认为是最强代码/推理模型之一",
+            "AWS Bedrock 和 Google Cloud 均为分发合作伙伴",
+            "2024 ARR 增速超 300%，约 $1.5B",
+        ],
+        "risks": [
+            "巨额算力成本，仍处亏损",
+            "OpenAI GPT-4o 和 Google Gemini Ultra 直接竞争",
+            "单一产品线风险",
+        ],
+    },
+    "Huawei": {
+        "name": "华为技术有限公司",
+        "valuation_usd": 220,
+        "rev_est": 99,
+        "rev_growth": 22,
+        "founded": 1987,
+        "founder": "任正非 (Ren Zhengfei)",
+        "hq": "深圳，广东，中国",
+        "employees": 207,
+        "segments": ["运营商业务（5G 设备）", "企业业务（云/存储/AI）", "终端业务（手机/PC/汽车）", "海思半导体（内部）"],
+        "comparables": ["ERIC", "NOK", "CSCO", "INTC"],
+        "ipo_status": "任正非多次公开表示不计划上市",
+        "last_funding": "员工持股计划，无外部融资",
+        "highlights": [
+            "2023 年营收 ~7042 亿人民币（约 $990 亿），同增 9.6%；2024 年超 $990 亿",
+            "麒麟 9000S 芯片突破 5G 封锁，Mate 60 系列引发广泛关注",
+            "鸿蒙 OS 生态持续扩大，2024 设备用户超 9 亿",
+            "全球 5G 专利数量第一，技术护城河深厚",
+        ],
+        "risks": [
+            "美国实体清单：高端芯片封锁持续",
+            "西方市场 5G 合同被排除（欧盟多国）",
+            "依赖自研替代进口，成本高企",
+        ],
+    },
+    "DJI": {
+        "name": "大疆创新 (DJI)",
+        "valuation_usd": 16,
+        "rev_est": 7,
+        "rev_growth": 15,
+        "founded": 2006,
+        "founder": "汪滔 (Frank Wang)",
+        "hq": "深圳，广东，中国",
+        "employees": 14,
+        "segments": ["消费级无人机（Mavic/Mini系列）", "行业级无人机（农业/测绘/影视）", "稳定器/相机云台（如影）", "农业植保机"],
+        "comparables": ["AVAV", "KTOS", "PRNT"],
+        "ipo_status": "2023 年港股 IPO 计划因美国制裁风险搁置",
+        "last_funding": "2021 年融资 $500M，估值 $16B",
+        "highlights": [
+            "全球消费无人机市场份额 >70%",
+            "Agras 农业无人机在中国市占率超 80%",
+            "2021 年被列入美国实体清单，仍是全球领导者",
+        ],
+        "risks": [
+            "美国国防部黑名单影响企业客户",
+            "本土竞争者（如道通、极飞）快速追赶",
+            "各国无人机监管趋严",
+        ],
+    },
+    "Stripe": {
+        "name": "Stripe",
+        "valuation_usd": 70,
+        "rev_est": 15,
+        "rev_growth": 25,
+        "founded": 2010,
+        "founder": "Patrick Collison, John Collison",
+        "hq": "San Francisco / Dublin",
+        "employees": 8,
+        "segments": ["支付处理", "Stripe Billing 订阅管理", "Stripe Capital 金融服务", "Stripe Atlas 公司注册"],
+        "comparables": ["V", "MA", "SQ", "PYPL", "ADYEN.AS"],
+        "ipo_status": "多次推迟；2024 年员工流动性计划估值 $65-70B；IPO 预计 2025-2026",
+        "last_funding": "2023 年 $6.5B 融资用于员工纳税，估值从峰值 $95B 降至 $50B 后回升",
+        "highlights": [
+            "2023 年处理支付额 ~$1 万亿，收入约 $15B",
+            "美国支付 API 市场开发者首选，生态极强",
+            "客户包括 Amazon、Shopify、Salesforce 等",
+        ],
+        "risks": [
+            "Adyen、PayPal、Square 直接竞争",
+            "利率上升增加支付成本",
+            "大客户谈判能力强，费率受压",
+        ],
+    },
+    "xAI": {
+        "name": "xAI (Elon Musk's AI company)",
+        "valuation_usd": 50,
+        "rev_est": 0.5,
+        "rev_growth": 500,
+        "founded": 2023,
+        "founder": "Elon Musk",
+        "hq": "Memphis, Tennessee, USA",
+        "employees": 1,
+        "segments": ["Grok AI 助手 (X 平台集成)", "Grok API", "超级计算机 Colossus（100k H100）"],
+        "comparables": ["MSFT", "GOOGL", "META"],
+        "ipo_status": "无上市计划",
+        "last_funding": "2025 年 $6B 融资，估值 $50B",
+        "highlights": [
+            "Grok 3 系列在多项 benchmark 上排名前列",
+            "Memphis Colossus 超级计算机为目前最大 AI 训练集群之一",
+            "与 X（Twitter）深度集成",
+        ],
+        "risks": [
+            "Elon Musk 精力分散（同时管理 Tesla/SpaceX/X/xAI）",
+            "起步晚，OpenAI/Google/Anthropic/Meta 先发优势显著",
+        ],
+    },
+}
+
+# Keywords that indicate the user is asking about their own portfolio/account
+_BROKER_INTENT_KW = (
+    "我的持仓", "我的仓位", "我持有", "我买了", "我的股票", "我的账户",
+    "账户余额", "可用余额", "我有多少钱", "我的资金", "我还有多少",
+    "持仓盈亏", "我的盈亏", "赚了多少", "亏了多少", "总盈亏", "当日盈亏",
+    "委托", "我的订单", "挂单", "我下的单",
+    "my portfolio", "my positions", "my holdings", "my account",
+    "my balance", "my orders", "how much cash", "what do i own",
+)
+
+
+def _is_broker_intent(message: str) -> bool:
+    """Return True if the message is asking about the user's own account/positions."""
+    low = message.lower()
+    return any(kw.lower() in low for kw in _BROKER_INTENT_KW)
+
+
+# ── Broker setup / configuration intent ──────────────────────────────────────
+# Maps broker type keys to their display names and recognition keywords.
+_BROKER_SETUP_NAMES: dict[str, tuple[str, tuple]] = {
+    "tiger":       ("老虎证券",   ("老虎", "tiger", "tigeropen", "tigerop")),
+    "futu":        ("富途牛牛",   ("富途", "futu", "futuooo", "牛牛", "futuo")),
+    "longbridge":  ("长桥证券",   ("长桥", "longbridge", "longport")),
+    "ibkr":        ("盈透证券",   ("盈透", "ibkr", "ib ", "interactive broker", "ib_insync")),
+    "alpaca":      ("Alpaca",     ("alpaca",)),
+    "webull":      ("Webull",     ("webull", "微牛")),
+    "xtquant":     ("迅投XTQuant", ("迅投", "xtquant", "qmt", "中信建投", "华鑫", "浙商")),
+    "easytrader":  ("EasyTrader", ("easytrader", "通达信", "同花顺", "ths", "tdx", "华泰", "国君")),
+}
+
+_BROKER_SETUP_VERBS = (
+    "配置", "连接", "接入", "安装", "设置", "开通", "使用", "启用", "绑定", "接口",
+    "添加券商", "添加交易", "怎么用", "如何用", "怎样用", "怎么接", "如何接",
+    "帮我", "帮忙", "教我", "指导", "setup", "connect", "configure", "install",
+    "add broker", "use broker", "how to use", "how to connect", "how to setup",
+    "guide", "tutorial", "凭证", "api key", "申请", "开发者",
+)
+
+_BROKER_GUIDE_INTENT_KW = (
+    "支持哪些券商", "有哪些券商", "券商列表", "券商能力", "能力矩阵",
+    "所有券商", "各个券商", "券商服务", "券商联动", "怎么连接各个券商",
+    "如何连接各个券商", "如何和各个券商连接", "怎么和各个券商连接",
+    "分析如何和各个券商连接", "使用这个项目的各个服务", "项目的各个服务",
+    "broker guide", "broker matrix", "supported brokers", "broker capabilities",
+    "broker services",
+)
+
+
+def _is_broker_guide_intent(message: str) -> bool:
+    """Return True for broad broker/service discovery requests.
+
+    These should show a capability guide instead of jumping straight into an
+    interactive broker-add wizard.
+    """
+    low = message.lower()
+    compact = low.replace(" ", "")
+    if any(kw in low or kw.replace(" ", "") in compact for kw in _BROKER_GUIDE_INTENT_KW):
+        return True
+    has_broker_word = any(k in low for k in ("券商", "证券", "broker"))
+    wants_overview = any(k in low for k in ("哪些", "列表", "支持", "能力", "矩阵", "各个", "所有", "服务"))
+    asks_connection = any(k in low for k in ("怎么", "如何", "怎样", "连接", "接入", "使用"))
+    return bool(has_broker_word and wants_overview and asks_connection)
+
+
+def _is_broker_setup_intent(message: str) -> bool:
+    """Return True if the user is asking to set up / configure a broker."""
+    low = message.lower()
+    has_verb = any(v in low for v in _BROKER_SETUP_VERBS)
+    if not has_verb:
+        return False
+    # Check if any broker name/keyword is mentioned
+    for _broker_type, (display, kws) in _BROKER_SETUP_NAMES.items():
+        if any(k in low for k in kws):
+            return True
+    # Generic broker-setup phrases without a specific broker named
+    generic_kws = ("券商", "证券", "broker", "交易所", "量化交易", "quant", "开户")
+    return any(k in low for k in generic_kws)
+
+
+def _detect_broker_type(message: str) -> str:
+    """Return the broker type key if a specific broker is mentioned, else ''."""
+    low = message.lower()
+    for broker_type, (_display, kws) in _BROKER_SETUP_NAMES.items():
+        if any(k in low for k in kws):
+            return broker_type
+    return ""
+
+
+_FINANCIAL_TERMS_BLOCKLIST: frozenset = frozenset({
+    "DCF", "EPS", "ROE", "ROA", "ROI", "PEG", "NAV", "AUM", "NMV", "IRR",
+    "NPV", "FCF", "OCF", "ICF", "EBIT", "WACC", "CAGR", "YTD", "QoQ",
+    "YoY", "MoM", "GDP", "CPI", "PPI", "PMI", "NFP", "ISM", "PCE",
+    "FED", "ECB", "PBC", "IMF", "BIS", "WTO", "SEC", "FSB",
+    "QE", "QT", "VIX", "RSI", "MACD", "EMA", "SMA", "ATR", "ADX",
+    "ATH", "ATL", "IPO", "SPO", "M&A", "LBO", "MBO",
+    "NYSE", "NASDAQ", "AMEX", "LSE", "TSX", "ASX", "SGX", "HKE",
+    "ETF", "CEF", "REIT", "MLP", "ADR", "GDR",
+    "USD", "CNY", "EUR", "GBP", "JPY", "HKD", "SGD", "AUD",
+    "BTC", "ETH", "XRP",  # crypto are handled by _CRYPTO_WORDS
+    "AI", "AR", "VR", "ML", "NLP", "API", "SaaS", "B2B", "B2C",
+})
+_FINANCIAL_TERMS_SYMBOL_BLOCKLIST: frozenset = frozenset(
+    str(item).upper() for item in _FINANCIAL_TERMS_BLOCKLIST
+)
+
+_MARKET_SYMBOL_WORD_BLOCKLIST: frozenset = frozenset({
+    "ANALYZE", "ANALYSIS", "DATA", "QUOTE", "QUOTES", "PRICE", "PRICES",
+    "MARKET", "MARKETS", "VOLUME", "VOLUMES", "CHART", "CHARTS", "PLOT",
+    "YAHOO", "STOOQ", "YFINANCE", "AKSHARE", "FINNHUB", "SOURCE",
+    "SOURCES", "PROVIDER", "PROVIDERS", "HISTORY", "HISTORICAL",
+})
+
+
+def _company_alias_symbol_blocklist() -> set[str]:
+    """Words like APPLE/TESLA are company aliases, not Yahoo ticker symbols."""
+    blocked: set[str] = set()
+    for name, ticker in _COMPANY_TO_TICKER.items():
+        alias = str(name or "").strip().upper()
+        resolved = str(ticker or "").strip().upper()
+        if (
+            1 < len(alias) <= 5
+            and alias.isalpha()
+            and resolved
+            and resolved not in ("未上市", "未独立")
+            and not resolved.startswith("PRIVATE:")
+            and alias != resolved
+        ):
+            blocked.add(alias)
+    return blocked
+
+
+_COMPANY_ALIAS_SYMBOL_BLOCKLIST: frozenset = frozenset(_company_alias_symbol_blocklist())
+
+
+def _is_blocked_market_symbol_candidate(symbol: str) -> bool:
+    """Return True for uppercase words that are context terms, not tickers."""
+    normalized = str(symbol or "").strip().upper()
+    if not normalized:
+        return True
+    return (
+        normalized in _FINANCIAL_TERMS_SYMBOL_BLOCKLIST
+        or normalized in _MARKET_SYMBOL_WORD_BLOCKLIST
+        or normalized in _COMPANY_ALIAS_SYMBOL_BLOCKLIST
+    )
+
+
+def _extract_market_symbol(message: str) -> str:
+    """Extract a likely market symbol from Chinese company names or tickers."""
+    resolved = _resolve_market_symbol(message)
+    if resolved:
+        return resolved
+    # 最长名称优先（避免"美的"抢先匹配"美的集团"）
+    for cn, tick in sorted(_COMPANY_TO_TICKER.items(), key=lambda x: -len(x[0])):
+        if cn in message:
+            return tick
+    # A股裸6位代码（600519 / sh600519 / 688256.SH）
+    m = _re_sym.search(r'(?<!\d)(?:[sS][hHzZ])?([036]\d{5}|68\d{4})(?:\.(?:SH|SZ|SS))?(?!\d)', message)
+    if m:
+        return m.group(1)
+    # Yahoo-compatible index / futures / FX tickers: ^GSPC, GC=F, EURUSD=X
+    m = _re_sym.search(r'(?<![A-Za-z0-9])(\^[A-Z0-9]{2,10}|[A-Z]{2,8}=F|[A-Z]{6}=X|DX-Y\.NYB)(?![A-Za-z0-9])', message, _re_sym.I)
+    if m:
+        return m.group(1).upper()
+    m = _re_sym.search(r'\b([A-Z]{1,5}(?:\.(?:HK|SH|SZ))?)\b', message)
+    if m and not _is_blocked_market_symbol_candidate(m.group(1)):
+        return m.group(1)
+    # Chinese text immediately after a ticker ("AAPL的市场") prevents \b from
+    # matching because Unicode word-boundary rules treat 的 as a word char.
+    m = _re_sym.search(r'(?<![A-Za-z])([A-Z]{1,5}(?:\.(?:HK|SH|SZ))?)(?![A-Za-z])', message)
+    if m and not _is_blocked_market_symbol_candidate(m.group(1)):
+        return m.group(1)
+    return ""
+
+def _extract_market_symbols(message: str, limit: int = 6) -> list[str]:
+    """Extract all likely market symbols, preserving mention order."""
+    hits: list[tuple[int, str]] = []
+
+    def add_hit(start: int, symbol: str) -> None:
+        if start < 0 or not symbol:
+            return
+        # Higher-priority resolvers win at the same text position. This avoids
+        # old aliases like "纳斯达克" -> QQQ duplicating the universe resolver's
+        # "纳斯达克" -> ^IXIC result.
+        if any(existing_start == start for existing_start, _ in hits):
+            return
+        hits.append((start, symbol))
+
+    for start, item in _resolve_market_mentions(message, limit=limit):
+        add_hit(start, item.symbol)
+
+    for name, ticker in sorted(_COMPANY_TO_TICKER.items(), key=lambda x: -len(x[0])):
+        if ticker.startswith("PRIVATE:") or ticker in ("未上市", "未独立"):
+            continue
+        start = message.find(name)
+        add_hit(start, ticker)
+
+    for match in _re_sym.finditer(r'(?<!\d)(?:[sS][hHzZ])?([036]\d{5}|68\d{4})(?:\.(?:SH|SZ|SS))?(?!\d)', message):
+        add_hit(match.start(), match.group(1))
+
+    for match in _re_sym.finditer(r'(?<![A-Za-z0-9])(\^[A-Z0-9]{2,10}|[A-Z]{2,8}=F|[A-Z]{6}=X|DX-Y\.NYB)(?![A-Za-z0-9])', message, _re_sym.I):
+        add_hit(match.start(), match.group(1).upper())
+
+    for match in _re_sym.finditer(r'\b([A-Z]{1,5}(?:\.(?:HK|SH|SZ))?)\b', message):
+        symbol = match.group(1)
+        if not _is_blocked_market_symbol_candidate(symbol):
+            add_hit(match.start(), symbol)
+
+    for match in _re_sym.finditer(r'(?<![A-Za-z])([A-Z]{1,5}(?:\.(?:HK|SH|SZ))?)(?![A-Za-z])', message):
+        symbol = match.group(1)
+        if not _is_blocked_market_symbol_candidate(symbol):
+            add_hit(match.start(), symbol)
+
+    ordered: list[str] = []
+    for _, symbol in sorted(hits, key=lambda item: item[0]):
+        symbol = str(symbol).upper()
+        if symbol not in ordered:
+            ordered.append(symbol)
+        if len(ordered) >= limit:
+            break
+    return ordered
+
+
+def _extract_symbol_from_history(history: list, max_lookback: int = 8) -> str:
+    """从最近的会话历史中提取标的（跟进问题继承上下文，如"现在的股票和趋势呢"）。
+
+    倒序扫描最近 max_lookback 条消息，返回最先命中的标的代码。
+    user 消息优先于 assistant 消息（用户提到的标的意图最明确）。
+    """
+    if not history:
+        return ""
+    recent = history[-max_lookback:]
+    # 先扫 user 消息（倒序）
+    for msg in reversed(recent):
+        if msg.get("role") != "user":
+            continue
+        content = msg.get("content", "")
+        if isinstance(content, list):   # 多模态消息取 text 部分
+            content = " ".join(p.get("text", "") for p in content
+                               if isinstance(p, dict) and p.get("type") == "text")
+        sym = _extract_market_symbol(content)
+        if sym:
+            return sym
+    # 再扫 assistant 消息（倒序）
+    for msg in reversed(recent):
+        if msg.get("role") != "assistant":
+            continue
+        sym = _extract_market_symbol(str(msg.get("content", "")))
+        if sym:
+            return sym
+    return ""
+
+def _is_stock_chart_analysis_request(message: str) -> bool:
+    """Return True for clear stock chart requests on stock-like symbols.
+
+    This intentionally catches both:
+    - "分析 AAPL 的 K 线图"
+    - "生成 Apple 公司近一年的股票图表"
+    so callers can route chart requests before generic market snapshot logic.
+    """
+    low = message.lower()
+    has_chart = any(k in low for k in (
+        "图表", "走势图", "k线", "k线图", "k-line", "kline", "chart", "plot",
+        "蜡烛图", "candlestick", "生成", "绘制", "画", "制作", "生成一个", "生成近",
+    ))
+    has_stock = any(k in low for k in ("股票", "stock", "股价")) or bool(_extract_market_symbol(message))
+    return has_chart and has_stock
+
+
+def _is_visual_market_artifact_request(message: str) -> bool:
+    """Return True for chart/dashboard/report requests tied to market data."""
+    low = message.lower()
+    has_visual = any(k in low for k in (
+        "图表", "走势图", "k线", "k线图", "k-line", "kline", "candlestick",
+        "chart", "plot", "dashboard", "看板", "晨报", "日报", "周报", "月报",
+        "report", "热力图", "heatmap",
+    ))
+    if not has_visual:
+        return False
+    has_market = any(k in low for k in (
+        "股票", "股价", "行情", "市场", "美股", "港股", "a股", "指数",
+        "持仓", "portfolio", "回测", "财报", "earnings", "基金", "etf",
+        "资产", "组合", "市场数据", "market data",
+    ))
+    return has_market or bool(_extract_market_symbol(message))
+
+
+_UNRESOLVED_CO_INDICATORS = (
+    # 中文公司后缀
+    "公司", "集团", "股份", "有限", "控股", "科技", "电子", "能源", "银行",
+    "汽车", "制药", "医疗", "传媒", "文化", "地产", "金融", "基金", "证券",
+    # 外国品牌/公司名常见关键词（未进入字典的）
+    "路易", "奔驰", "宝马", "大众", "爱马仕", "香奈儿", "耐克", "阿迪",
+    "三星", "索尼", "丰田", "本田", "日产", "空客", "波音", "壳牌", "埃克森",
+    "高盛", "摩根", "花旗", "巴克莱", "德银", "瑞信", "瑞银",
+    "奈飞", "推特", "脸书", "优步", "滴滴", "字节", "快手", "拼多多",
+)
+
+
+def _has_unresolved_company_mention(message: str) -> bool:
+    """Return True if message names a company/brand that _extract_market_symbol couldn't resolve.
+
+    Used to block silent history inheritance: if user says "路易斯威登怎么看" and we
+    can't resolve it, we should NOT fall back to the last queried ticker (e.g. 300124).
+    """
+    if _extract_market_symbol(message):
+        return False  # already resolved — no problem
+    return any(kw in message for kw in _UNRESOLVED_CO_INDICATORS) or _looks_like_unresolved_market_name(message)
+
+
+# These keywords signal the user is asking about property markets, NOT stocks.
+# Any message matching here must NOT be routed to the stock market snapshot path
+# and must NOT silently inherit a stock ticker from session history.
+_REALTY_QUERY_KEYWORDS = (
+    "房价", "楼市", "楼价", "房产", "房地产", "地产", "租金", "租房",
+    "二手房", "新房", "住宅", "商品房", "开发商", "楼盘", "物业",
+    "house price", "property price", "real estate market", "housing market",
+    "home price", "apartment rent",
+)
+# Major Chinese cities — when paired with real-estate context words, confirms realty query.
+_CN_CITIES = (
+    # 一线
+    "北京", "上海", "深圳", "广州",
+    # 新一线
+    "成都", "杭州", "武汉", "南京", "重庆", "西安", "长沙", "天津",
+    "苏州", "郑州", "东莞", "厦门", "宁波", "青岛", "无锡", "合肥",
+    "佛山", "福州",
+    # 二线
+    "南宁", "昆明", "贵阳", "沈阳", "大连", "哈尔滨", "长春", "济南",
+    "石家庄", "太原", "呼和浩特", "南昌", "兰州", "银川", "西宁",
+    "乌鲁木齐", "海口", "三亚", "珠海", "中山", "惠州", "温州", "南通",
+    "嘉兴", "金华", "绍兴", "泉州", "漳州", "赣州", "九江",
+    "烟台", "潍坊", "济宁", "临沂", "洛阳", "南阳", "开封",
+    "湘潭", "株洲", "常德", "常州", "徐州", "扬州", "盐城",
+    "唐山", "保定", "邯郸", "秦皇岛", "廊坊",
+    "湛江", "江门", "茂名", "汕头", "汕尾",
+    "桂林", "柳州", "芜湖", "马鞍山", "蚌埠",
+)
+
+# International city keywords (for realty query detection — no akshare data but LLM context)
+_INTL_CITIES = (
+    # English names
+    "new york", "los angeles", "san francisco", "chicago", "miami",
+    "london", "paris", "berlin", "amsterdam", "zurich", "geneva",
+    "tokyo", "osaka", "seoul", "singapore", "sydney", "melbourne",
+    "hong kong", "taipei", "bangkok", "dubai", "toronto", "vancouver",
+    # Chinese transliterations / common references
+    "纽约", "洛杉矶", "旧金山", "芝加哥", "迈阿密",
+    "伦敦", "巴黎", "柏林", "阿姆斯特丹", "苏黎世",
+    "东京", "大阪", "首尔", "新加坡", "悉尼", "墨尔本",
+    "香港", "台北", "曼谷", "迪拜", "多伦多", "温哥华",
+)
+# Market/analysis words that only count as "stock" context — not real-estate context
+_STOCK_ONLY_MARKET_WORDS = (
+    "股票", "股价", "涨跌", "涨幅", "股市", "A股", "港股", "美股",
+    "指数", "大盘", "个股", "板块", "开盘", "收盘", "换手率",
+    "stock", "share", "equity", "ticker",
+)
+
+
+def _is_realty_query(message: str) -> bool:
+    """Return True when the message is about real-estate / housing, not stocks.
+
+    Three patterns count:
+      (a) explicit realty keywords (房价, 楼市, 租金 …)
+      (b) CN city name + generic market/price/trend words with no stock-specific keywords
+      (c) international city name + realty keyword
+    """
+    low = message.lower()
+    # (a) Explicit realty keyword present
+    if any(k in low for k in _REALTY_QUERY_KEYWORDS):
+        return True
+    # (b) CN city + market/trend word but NO stock-specific word
+    has_cn_city = any(c in message for c in _CN_CITIES)
+    has_market_word = any(k in low for k in ("市场", "价格", "走势", "趋势", "现在", "最新"))
+    has_stock_word  = any(k in low for k in _STOCK_ONLY_MARKET_WORDS)
+    if has_cn_city and has_market_word and not has_stock_word:
+        return True
+    # (c) International city + realty/housing word
+    has_intl_city = any(c in low for c in _INTL_CITIES)
+    if has_intl_city and has_market_word and not has_stock_word:
+        return True
+    return False
+
+
+def _detect_market_overview(message: str):
+    """Detect a whole-market overview request → 'cn' | 'us' | 'hk' | None.
+
+    Critical disambiguation: "分析A股市场行情" means the A-share MARKET (indices),
+    NOT a stock with ticker 'A' (Agilent). This must run before any ticker
+    extraction so a market name never collapses into a wrong single stock.
+    Returns None for coding/chart/realty requests or specific-stock queries.
+    """
+    if _is_realty_query(message):
+        return None
+    low = message.lower().strip()
+    _coding_kw = ("write", "generate", "script", "code", "plot", "backtest",
+                  "策略", "代码", "回测", "编写", "k线", "kline", "python", "dashboard")
+    if any(k in low for k in _coding_kw):
+        return None
+    if _is_stock_chart_analysis_request(message) or _is_visual_market_artifact_request(message):
+        return None
+
+    # Market-level intent words (overview, not a single stock)
+    _overview_kw = (
+        "市场", "大盘", "行情", "今天怎么样", "今日", "现在怎么样", "怎么样",
+        "收盘", "开盘", "走势", "整体", "表现", "概况", "overview", "market",
+    )
+    if not any(k in low for k in _overview_kw):
+        return None
+
+    # Market identifiers (ordered: check before generic '大盘' defaults to CN)
+    _us = ("美股", "美国股市", "美国市场", "华尔街", "纳斯达克", "纳指", "标普",
+           "道琼斯", "道指", "us market", "wall street", "nasdaq", "s&p", "dow")
+    _hk = ("港股", "恒生", "香港股市", "香港市场", "hk market", "hang seng", "hsi")
+    _cn = ("a股", "a 股", "沪深", "上证", "深证", "创业板", "沪市", "深市",
+           "科创板", "a-share", "ashare", "china market", "中国股市", "大盘")
+
+    if any(k in low for k in _us):
+        return "us"
+    if any(k in low for k in _hk):
+        return "hk"
+    if any(k in low for k in _cn):
+        return "cn"
+    return None
+
+
+def _is_market_snapshot_request(message: str, history: list = None) -> bool:
+    """Return True for simple quote / market snapshot analysis requests.
+
+    跟进问题（"现在的股票和趋势呢"）：当前消息无标的时回溯会话历史继承。
+    但如果消息中包含未能识别的公司名，则不视为跟进问题（不继承历史标的）。
+    房价/楼市类问题直接拒绝，防止继承上一个股票代码。
+    """
+    # Real-estate questions must never be routed to the stock snapshot path
+    if _is_realty_query(message):
+        return False
+    # Pre-built prompts from cmd_analyze contain specific footer text — bypass snapshot
+    if ("Please provide a comprehensive analysis of" in message
+            or "请对以上" in message and "进行综合分析" in message):
+        return False
+    # Inline check (avoids circular import with aria_cli.py)
+    _coding_kw = ("write", "generate", "create", "script", "code", "plot",
+                  "backtest", "策略", "代码", "回测", "编写", "生成", "k线",
+                  "k-line", "kline", "python", "dashboard", "写一个", "写代码")
+    _is_coding = any(k in message.lower() for k in _coding_kw)
+    if _is_coding or _is_stock_chart_analysis_request(message) or _is_visual_market_artifact_request(message):
+        return False
+    low = message.lower()
+    has_market_word = any(k in low for k in (
+        "市场", "行情", "股价", "价格", "涨跌", "涨幅", "现价", "今天",
+        "现在", "最新", "分析", "走势", "趋势", "market", "quote", "price",
+    ))
+    if not has_market_word:
+        return False
+    if _extract_market_symbol(message):
+        return True
+    # Only inherit history if the message doesn't name an unresolvable company
+    if _has_unresolved_company_mention(message):
+        return False
+    return bool(history and _extract_symbol_from_history(history))
+
+
+def _format_compact_market_cap(value, currency: str = "USD") -> str:
+    try:
+        cap = float(value)
+        if cap <= 0:
+            return "—"
+        if cap >= 1e12:
+            return f"{currency} {cap / 1e12:.2f}T"
+        if cap >= 1e9:
+            return f"{currency} {cap / 1e9:.1f}B"
+        if cap >= 1e6:
+            return f"{currency} {cap / 1e6:.0f}M"
+        return f"{currency} {cap:,.0f}"
+    except Exception:
+        return "—"
+
+
+def _market_snapshot_trend(price, high, low, change_pct) -> str:
+    try:
+        p = float(price)
+        h = float(high) if high not in (None, "", "—") else None
+        l = float(low) if low not in (None, "", "—") else None
+        c = float(change_pct) if change_pct not in (None, "", "—") else 0.0
+        parts = []
+        if h is not None and l is not None and h > l:
+            pos = int((p - l) / (h - l) * 100)
+            if pos >= 75:
+                parts.append("near day high")
+            elif pos <= 25:
+                parts.append("near day low")
+            else:
+                parts.append("mid-range")
+        if abs(c) >= 2:
+            parts.append("strong up" if c > 0 else "strong down")
+        elif c != 0:
+            parts.append("up" if c > 0 else "down")
+        return ", ".join(parts) or "flat"
+    except Exception:
+        return "—"