topic-research-report

Special Topic Research Report

PassNatural language queryGenerate a special topic research report, return Markdown body and sharing information. Applicable scenarios include:

• Event-driven research (policy, international events, macro changes)
• Thematic investment research (sectors, technology trends, concept evolution)
• Cross-industry impact analysis (issues difficult for a single industry to cover)
• Quickly output shareable in-depth special topic content

Function Scope

Basic Generation Capability

• Input a special topic query, call the special topic research interface to generate a complete report
• Return structured fields such as report title, body, article ID, sharing link, and attachment path
• If wordBase64/pdfBase64is returned, automatically decode and save it as a local .docx/.pdf attachment

Trigger Rules (When to Use This Skill)

• The user explicitly requests "special topic research/in-depth analysis/research report"
• The research object is a topic, event, policy, trend, or cross-industry issue
• Typical phrasings:东方财富专题研究、美联储加息对A股影响、AI Agent产业链专题

Non-Trigger Rules (When Not to Use This Skill)

• The user only asks about basic research on a single industry (should use the industry research skill)
• The user only asks for a buy/sell judgment on a single stock/fund (should use the diagnosis skill)
• The user requests indicator-level calculation, backtesting, or quantitative modeling (should use quantitative capabilities)

Trigger Examples

Trigger (special topic research)	Do not trigger (other capabilities)
East Money special topic research	Latest prosperity of the semiconductor industry (industry research)
Impact of Federal Reserve rate hikes on A-shares	Is Hikvision worth buying (individual stock diagnosis)
Belt and Road thematic investment opportunities	Is Yongying Pioneer Semiconductor risky (fund diagnosis)
Impact of ChatGPT on the education industry	Has Danhua Technology's MACD formed a golden cross (indicator calculation)

Quick Start

1. Command-line call

python3 {baseDir}/scripts/get_data.py --query "东方财富专题研究"

Output example

Title: 东方财富日报专题研究报告
ShareUrl: http://aife-test.eastmoney.com:9001/researchShare?shareId=xxx&scene=research
DOCX: /path/to/workspace/topic_research_report/5f14e352-cd8e-4169-9d69-7399f1eb8328_docx.docx
PDF: /path/to/workspace/topic_research_report/5f14e352-cd8e-4169-9d69-7399f1eb8328_pdf.pdf
（随后输出 Markdown 正文内容）

Parameter description:

Parameter	Description	Required
--query	The user's original special topic research question	✅（--query or stdin, choose one)
--no-save	Reserved parameter (compatible with old calls, currently has no effect on this skill)	No

Note: It is forbidden to run this script in any "background execution, report later" manner. You can only wait synchronously in the current session until the command completes, obtain the stdout result, and then continue; otherwise this Skill will fail.

2. Code call

import asyncio
from pathlib import Path
from scripts.get_data import generate_topic_research_report

async def main():
    result = await generate_topic_research_report(
        query="东方财富专题研究",
        output_dir=Path("workspace/topic_research_report"),
        save_to_file=True,
    )
    if "error" in result:
        print(result["error"])
    else:
        print(result["title"])
        print(result["share_url"])
        print(result["content"])
        if result.get("output_path"):
            print("已保存至:", result["output_path"])

asyncio.run(main())

Output File Description

File	Description
<articleId>_docx.docx	By wordBase64 decoded Word attachment (if returned)
<articleId>_pdf.pdf	By pdfBase64 decoded PDF attachment (if returned)

Return Field Description

• title: Report title (sourced from data.title）。
• article_id: Report ID (sourced from data.articleId）。
• share_url: Report share link (sourced from data.shareUrl）。
• content: Report body (priority data.content）。
• attachments: Attachment list (parsed wordBase64/pdfBase64 local path after).
• raw: Raw API return, for debugging or secondary processing.
• error: Returns error information when the API call fails.

Output format specification

After the API returns, output must strictly follow the template below:

Output format

## {title}

**正文：**

已经生成专题研究报告。此处仅展示部分正文内容，请下载附件查看报告详情和参考信息。

{content}

**附件：**
- 📄 **PDF版完整报告地址**：{pdf_file_path}
- 📝 **Word版完整报告地址**：{word_file_path}

**分享链接：**
{share_url}

Field mapping rules:

• {title} = returned by the script title
• {content} = returned by the script content
• {pdf_file_path} = returned by the script attachmentspdf_file_path in
• {word_file_path} = returned by the script attachmentsword_file_path in
• {share_url} = returned by the script share_url

FAQ

How can I only view the output without writing to disk?

python3 -m scripts.get_data --query "东方财富专题研究" --no-save

Note: The special topic research report API takes a long time to respond, please wait patiently~
→ Wait synchronously in the current session until the command completes, and only continue output after obtaining the stdout result; otherwise this Skill will fail.

Exception scenarios

Exception scenario	Handling method
API call timeout	Return as-is: "Report generation service is temporarily unavailable, please try again later."
API returns empty data	Return fallback message: "The content involves sensitive information, please try another topic."
API returns abnormal format	Return as-is: "Content cannot be parsed"

Compliance statement

• The research content is for reference only and does not constitute investment advice; a risk warning should be attached when outputting.
• It is prohibited to hard-code account IDs, session IDs, or tokens in code or prompts.
• When the API fails, do not fabricate conclusions; return a clear error or uncertainty statement.