针对web3 的网站上的新闻进行拉取， 训练 GPT， 并能够给出具有权重对比的新闻

[iannono]

我们希望能够增加一些我们常看的网站， GPT 能够通过阅读这些网站的新闻， 进行过滤， 给出值得关注的内容， 我们不需要简单的摘要， 而是希望有训练的过程在里面， 能够帮助我们去筛选出重要的新闻， 或者新闻间的关联关系进行推理

[0xRichardH]

https://news-knowledge-base.tonob.net/

[0xRichardH]

https://chatgpt.com/g/g-uxiFxXzjc-better-news

{
  "openapi": "3.1.0",
  "info": {
    "title": "PA News Knowledge Base",
    "description": "API to retrieve PA News Spider data.",
    "version": "1.0.0"
  },
  "servers": [
    {
      "url": "https://news-knowledge-base.tonob.net",
      "description": "Main (production) server"
    }
  ],
  "paths": {
    "/PANewsSpider__{date}.json": {
      "get": {
        "operationId": "getPANewsSpiderData",
        "summary": "Retrieves PA News Spider data for a specified date.",
        "description": "Fetches the PA News Spider data available in JSON format for the specified date.",
        "parameters": [
          {
            "name": "date",
            "in": "path",
            "required": true,
            "description": "The date for which to retrieve the PA News Spider data, formatted as YYYYMMDD.",
            "schema": {
              "type": "string",
              "pattern": "^\\d{8}$"
            }
          }
        ],
        "responses": {
          "200": {
            "description": "Successful response with PA News Spider data.",
            "content": {
              "application/json": {
                "schema": {
                  "type": "object",
                  "properties": {
                    "exampleProperty": {
                      "type": "string"
                    }
                  }
                }
              }
            }
          }
        }
      }
    }
  }
}

---

Issue

{
  "response_data": "ResponseTooLargeError"
}

[0xRichardH]

https://github.com/ginobefun/BestBlogs/blob/main/flows/Dify/ArticleAnalysisFlow.md

https://github.com/ginobefun/BestBlogs/blob/main/flows/Dify/BestBlogs.dev%20%E5%9F%BA%E4%BA%8E%20Dify%20Workflow%20%E7%9A%84%E6%96%87%E7%AB%A0%E6%99%BA%E8%83%BD%E5%88%86%E6%9E%90%E5%AE%9E%E8%B7%B5.md