Market Daily Review

WorkBuddy PandaData 数据源覆盖

本技能在 WorkBuddy 专家包中运行时，所有实时和历史金融数据必须来自已连接的 pandadata Connector。若下文、参考资料或脚本提到 Python panda_data SDK、 AkShare、Tushare、网页抓取、直连 HTTP 或本地凭证，以本节为准：不得使用这些 方式获取正式数据。

1. 数据型任务必须先完成 auth_status 和至少一次真实的 call_pandadata 业务调用；在收到 Connector 返回前，禁止输出分析、排名、数字结论或“无数据”。本 Skill 的流程不得绕过 主 Agent 的最低接口调用清单。
2. 先按本 Skill 的 references/pandadata-interface-contracts.md 选择已登记方法和参数。只要任务能映射到已登记接口， 就直接通过 call_pandadata 传入该业务方法和 params；常规调用前不得执行接口检索。
3. 仅当已登记方法被 Connector 明确报告为参数契约不兼容、字段契约变化或调用失败时， 才对该方法调用一次 get_method_doc，修正参数后最多重试一次。
4. 仅当本地接口表没有匹配项，或 Connector 明确报告方法不存在/不受支持时，才调用 search_methods 动态发现接口；不得靠猜测连续试用名称相近的 get_* 方法。
5. 只有 call_pandadata 实际返回 0 行时才允许写“无数据”。必须先完成一次复查调用：校验 最新交易日与代码格式，放宽日期窗口，移除非必填过滤条件，或使用登记的备用参数；仍为 0 行才可如实报告，并保留两次调用回执。0 行不得触发 search_methods。
6. 不向 call_pandadata 添加未登记参数或顶层行数限制；记录实际方法、参数、数据日期、 频率、复权口径、行数、空值和错误状态。
7. 包内脚本只可处理 Connector 已返回的数据或执行纯本地计算与校验，不得自行联网取数。

最终答案必须包含“数据调用回执”表：接口、实际参数、状态、行数、数据日期范围和关键字段。 缺少回执表示任务未完成，必须继续调用工具而不是结束回答。

权限不足、配额限制、空结果、延迟发布和字段缺失都必须明确披露，不得切换到其他数据源 或用模型推断补数。

{
  "version": 1,
  "task": {
    "placeholder": "补充复盘口径或特别要求（可选）",
    "required": false
  },
  "fields": [
    {
      "key": "date",
      "label": "复盘日期",
      "type": "date",
      "help": "留空时使用最近已完成的 A 股交易日"
    },
    {
      "key": "focus",
      "label": "重点关注",
      "type": "text",
      "placeholder": "例如：北向持仓、半导体板块、龙虎榜"
    }
  ],
  "prompt_template": "{{#task}}任务与材料：\n{{task}}\n\n{{/task}}{{#attachments}}用户上传的材料（已放入工作区）：\n{{attachments}}\n\n{{/attachments}}执行 A 股收盘复盘。{{#date}}以 {{date}} 为目标日期。{{/date}}未指定日期时以最近已完成的 A 股交易日为目标日期。{{#focus}}重点关注 {{focus}}。{{/focus}}先核验交易日，再覆盖指数表现与估值、市场宽度、涨跌停情绪、行业与概念、龙虎榜、大宗交易、两融、北向持仓和风险提示，逐项标注数据接口与日期并说明滞后或缺失数据，输出中文报告。"
}

Market Daily Review

Use this skill to generate factual A-share after-close review reports. Prefer Pandadata as the data source, keep every statistic traceable to an interface and data date, and never invent missing figures.

Workflow

1. Determine the target date. If the user does not provide one, use the latest completed A-share trading day. Check get_last_trade_date and get_trade_cal; if the target date is closed, return a short "今日休市" note instead of a full report.
2. Load pandadata before making real API calls. Use its method index or search script to confirm parameters and fields; do not guess Pandadata signatures.
3. Collect data in this order:
   • Trading calendar and stock universe: get_last_trade_date, get_trade_cal, get_trade_list.
   • Index performance and valuation: get_index_daily, get_index_indicator.
   • Market breadth and sentiment: get_stock_daily or get_stock_rt_daily, plus get_stock_status_change.
   • Hot sectors and concepts: get_industry_constituents, get_concept_list, get_concept_constituents.
   • Funds and notable trades: get_lhb_list, get_lhb_detail, get_block_trade, get_margin, get_hsgt_hold.
4. Compute breadth and ranking metrics from raw rows: rising/falling counts, limit-up/limit-down counts, turnover leaders, industry/concept leaders, 龙虎榜 top net buy/sell names, block-trade discount/premium distribution, margin balance change, and northbound holding changes.
5. Generate Markdown using references/report-template.md. Save the report to reports/daily/YYYYMMDD.md unless the user gives another path.
6. Run scripts/validate_report.py <report-path> after writing the report. Fix missing sections, missing source notes, or missing data-date labels before presenting the result.

Pandadata Reference

Read references/pandadata-map.md when planning calls, selecting fields, or deciding how to degrade if a data interface is unavailable. The map is a routing aid only; the exact call contract must still come from pandadata.

Report Rules

• Write in Chinese unless the user requests another language.
• Use absolute dates such as 2026-06-11; avoid ambiguous "today" in the final report body.
• Mark T+1 datasets clearly. Margin financing, northbound holdings, and some exchange disclosures may lag the market date.
• State the limit-up/limit-down counting rules, including whether ST stocks and one-price boards are included.
• Keep the report factual: summarize structure, flows, and anomalies; do not give tomorrow's trading instructions or personalized investment advice.
• When a data call fails, keep the report useful by generating available sections and adding a concise missing-data note under "数据说明".

Automation

When the user asks for an automated daily review, create an after-close task for trading days only, preferably after 18:30 Asia/Shanghai so delayed datasets have time to settle. Make the task idempotent: if reports/daily/YYYYMMDD.md already exists, regenerate and overwrite it.