daily-tech-news

Daily Tech News - In-depth Daily Tech News Collection

Automatically search for and summarize the day's tech news, usingmulti-source cross-validation + in-depth research + importance scoringmechanism to generate high-quality Markdown blog posts.

⚠️ Required parameter

When calling this skill, the target date must be specified in the prompt:

TARGET_DATE: YYYY-MM-DD 格式的日期
示例: 2026-02-02

Correct way to call:

✅ "Run /daily-tech-news to collect tech news for 2026-02-02"
✅ "Run the daily-tech-news skill, date: 2026-02-02"
❌ "Run /daily-tech-news" (missing date parameter)

All dates during execution (file names, frontmatter, titles, search keywords) must use this TARGET_DATE.

Core Principles

Accuracy first → multi-source cross-validation + in-depth reading of original articles Quality over quantity → importance scoring filtering + deduplication and noise reduction Depth over breadth → in-depth analysis of the Top 3 news items + technical interpretation

Phase 1: Preparation Stage

Step 1.1: Determine the search date

Important: The date must be obtained from the calling parameters, not calculated on your own.

When GitHub Actions or a user calls this skill, the target date will be specified in the prompt. You must:

Extract the date from the prompt: Find the date explicitly specified in the prompt (format: YYYY-MM-DD)
Use that date as the TARGET_DATE: All subsequent operations are based on this date
Do not calculate the date yourself: Do not use relative dates such as "yesterday" or "today"

TARGET_DATE = 从 prompt 参数中提取的日期
格式: YYYY-MM-DD
示例: 2026-02-02

Date usage rules:

File name:daily-tech-news-{TARGET_DATE}.mdx
frontmatter date：{TARGET_DATE}
Title date: Convert TARGET_DATE to Chinese format (e.g., 2026年02月02日)
Search keywords: Use TARGET_DATE

Step 1.2: Initialize the deduplication cache

Check news published in the past 7 days to avoid duplicate coverage:

读取 content/blog/daily-tech-news-*.mdx 最近 7 篇
提取已报道的：项目名、公司名、事件关键词
建立去重列表

Phase 2: Multi-source In-depth Search

Step 2.1: Layer 1 - Authoritative official sources (highest priority)

Authoritative sources that must be visited:

Source	URL	Description
Hacker News	https://news.ycombinator.com/	Tech community bellwether
GitHub Trending	https://github.com/trending	Open-source project trends
Trendshift	https://trendshift.io/	GitHub trend analysis
Product Hunt	https://www.producthunt.com/	New product releases

Use WebFetch to directly retrieve, and extract the day's popular content (Top 10).

Step 2.2: Layer 2 - Professional media sources

Search 2-3 professional keyword combinations for each field:

AI / LLM field

搜索策略（按优先级）：
1. site:techcrunch.com AI {date}
2. site:theverge.com artificial intelligence {date}
3. "OpenAI" OR "Anthropic" OR "Google AI" announcement {date}
4. "LLM" OR "GPT" OR "Claude" release {date}
5. machine learning breakthrough research {date}

GitHub / open-source field

搜索策略：
1. site:github.blog {date}
2. "open source" major release {date}
3. GitHub "stars" trending repository {date}

Frontend development

搜索策略：
1. site:reactjs.org OR site:vuejs.org blog {date}
2. "React" OR "Vue" OR "Next.js" OR "Svelte" release {date}
3. frontend framework update {date}
4. JavaScript TypeScript major update {date}

Backend / infrastructure

搜索策略：
1. site:kubernetes.io blog {date}
2. "Rust" OR "Go" programming release {date}
3. "Docker" OR "Kubernetes" announcement {date}
4. cloud infrastructure AWS Azure GCP {date}

Tech industry trends

搜索策略：
1. site:crunchbase.com funding {date}
2. tech startup Series A B C funding {date}
3. developer tools company announcement {date}

Step 2.3: Layer 3 - Chinese tech communities

Supplement with original Chinese content:

搜索策略：
1. site:juejin.cn 热门 {date}
2. site:infoq.cn {date}
3. site:segmentfault.com 头条 {date}

Phase 3: In-depth Research and Verification

Step 3.1: Information cross-validation

For each candidate news item, perform:

1. 核心事实提取：
   - WHO: 涉及的公司/人物/项目
   - WHAT: 具体发生了什么
   - WHEN: 准确时间
   - WHERE: 发生地点/平台
   - WHY: 为什么重要

2. 多源验证（至少 2 个独立来源）：
   - 来源 A 报道 → WebFetch 获取原文
   - 来源 B 验证 → WebSearch 搜索相同事件
   - 对比核心事实是否一致

3. 置信度标记：
   ✓ 已验证（2+ 源确认）
   ? 待验证（仅单一来源）
   ⚠️ 有争议（来源间有矛盾）

Step 3.2: In-depth reading of original articles

For the Top 10 candidate news items, use WebFetch to retrieve the full content:

WebFetch 提取要点：
1. 文章核心论点
2. 关键数据和事实
3. 专家引用和评论
4. 技术细节和实现方式
5. 对开发者的实际影响

Step 3.3: Importance Scoring

Score each news item (out of 100 points):

Dimension	Weight	Scoring Criteria
Scope of Impact	25%	How many developers does it affect? Global vs local
Practicality	25%	Can developers apply it immediately? What problem does it solve?
Novelty	20%	First announcement vs repeated coverage? Breakthrough vs incremental?
Depth of Information	15%	Technical details vs title only?
Authority	15%	Official release vs hearsay? How reliable is the source?

Scoring example:

Kubernetes 1.33 原生支持 Sidecar
- 影响范围: 24/25 (全球 K8s 用户)
- 实用性: 23/25 (直接改变部署方式)
- 新颖性: 18/20 (首次正式发布)
- 信息深度: 14/15 (有技术实现细节)
- 权威性: 15/15 (官方发布)
= 总分: 94/100 ⭐⭐⭐⭐⭐

Step 3.4: Deduplication and Noise Reduction

Exclude the following content:

❌ 标题党检测：
- 标题含"震惊"、"重磅"、"曝光"但无实质内容
- 标题与正文关键词严重不匹配
- 过多感叹号和问号

❌ 低质内容过滤：
- 与开发者无关的泛科技新闻
- 无具体信息的笼统报道（如"AI 继续发展"）
- 7 天内已报道过的重复事件
- 单一来源且无法验证的消息

❌ 内容去重：
- 同一事件的多篇报道只保留信息最丰富的一篇
- 优先保留官方来源 > 专业媒体 > 泛媒体

Phase 4: Top 3 In-Depth Analysis

Step 4.1: Select the 3 most important news items of the day

Based on the importance scores, select the 3 news items with the highest scores for in-depth analysis.

Step 4.2: In-Depth Analysis Template

For each Top 3 news item, write an in-depth analysis:

### [标题] ⭐⭐⭐⭐⭐

**核心要点：**
- 要点1：[具体事实]
- 要点2：[具体事实]
- 要点3：[具体事实]

**技术解读：**
[这意味着什么？对开发者有什么影响？技术上如何实现？]

**背景上下文：**
[这个事件的历史背景是什么？为什么现在发生？]

**开发者行动建议：**
- 建议1：[具体可执行的行动]
- 建议2：[具体可执行的行动]

**相关链接：**
- 官方公告: [链接]
- 技术文档: [链接]
- 社区讨论: [链接]

Phase 5: Generate Blog Post

Step 5.1: File Structure

In content/blog/ directory, create an MDX file.

Important: must use the TARGET_DATE

filename: daily-tech-news-{TARGET_DATE}.mdx

Step 5.2: Content Template

Date format conversion:

TARGET_DATE format:2026-02-02
Chinese date format:2026年02月02日
frontmatter date format:2026-02-02(same as TARGET_DATE)

---
title: 每日技术资讯 - {TARGET_DATE 转换为中文格式}
description: {当日资讯亮点摘要，50-100字，突出 Top 3。末尾添加"今日共收录N条经多源验证的重要技术资讯。"}
date: {TARGET_DATE}
---

## 🔥 今日焦点

{Top 3 资讯的深度分析，使用 Phase 4 的深度分析模板}

---

## AI / 人工智能

### {标题} ⭐⭐⭐⭐

{摘要内容，3-5句话，包含核心事实}

**为什么重要：** {一句话解释对开发者的影响}

- 来源: [{来源名称}]({URL})
- 验证: ✓ 多源确认

### {标题2}
...

## GitHub / 开源

### GitHub 热门项目

本日 GitHub 趋势榜热门项目：

- **[owner/repo-name](https://github.com/owner/repo-name)** (TypeScript, 12.5k ⭐) ⭐⭐⭐⭐
  AI 驱动的代码编辑器插件，支持多种 IDE。
  **亮点：** {为什么这个项目值得关注}

- **[owner/another-repo](https://github.com/owner/another-repo)** (Python, 8.2k ⭐) ⭐⭐⭐
  轻量级机器学习框架，专注于边缘设备部署。

- 来源: [GitHub Trending](https://github.com/trending), [Trendshift](https://trendshift.io/)

### {其他开源资讯}
...

## 前端开发

...

## 后端 / 基础设施

...

## 科技动态

...

---

## 📊 今日数据

| 指标 | 数值 |
|------|------|
| 搜索源数量 | {N} 个 |
| 候选资讯 | {N} 条 |
| 去重后 | {N} 条 |
| 最终收录 | {N} 条 |
| 多源验证率 | {N}% |

---

> 本文由 Claude 自动生成，采用多源交叉验证机制。如发现错误，欢迎反馈。

Phase 6: Quality Check and Submission

Step 6.1: Final Quality Check

□ 所有资讯都有来源链接
□ Top 3 资讯都有深度分析
□ 重要性评分标记完整
□ 无重复内容
□ 无明显事实错误
□ MDX 安全检查（必须逐项确认）：
  □ 正文中无裸露的 < 符号（代码块内除外），已用 &lt; 替代
  □ 代码块语言标识符均为 Shiki 支持的语言（不确定则用 text）
  □ 正文中无裸露的 {} 花括号（代码块内除外）
□ 中文表达通顺

Step 6.2: Submit Changes

git add content/blog/daily-tech-news-{date}.mdx
git commit -m "docs: add daily tech news for {date}"

Output Requirements

Requirement	Description
Language	All in Chinese
Format	MDX file, compatible with Next.js blog
Regular news length	50-150 word summary
Top 3 in-depth analysis	200-400 words per item
Source attribution	Must indicate original link + verification status
Number of categories	At least 3 categories
Total number of news items	10-20 items (quality first)
Importance score	Each news item marked 1-5 stars

Information source priority

First priority (authoritative official)

GitHub official blog, Trending
Official blogs of various frameworks/languages (React, Vue, Rust, Go, etc.)
Official company announcements (Google, Microsoft, AWS, etc.)

Second priority (professional media)

Hacker News (community bellwether)
TechCrunch, The Verge (in-depth tech reporting)
InfoQ, Juejin (Chinese tech communities)

Third priority (general media)

General tech news websites
Social media hot topics
Personal blogs

Clickbait identification rules

Automatically downweight or exclude:

❌ 情绪化词汇过多：震惊、重磅、曝光、惊天、颠覆
❌ 过度夸张：史上最强、彻底改变、完全碾压
❌ 模糊表述：某公司、据说、可能、或将
❌ 无实质内容：AI 继续发展、技术在进步
❌ 标题与内容不符：标题说重大突破，内容只是小更新

⚠️ MDX Safety Rules (Must Be Strictly Followed)

MDX is not ordinary Markdown; it parses content as JSX. The following rulesmustbe followed when generating content, otherwise it will cause the Vercel build to fail:

Rule 1: Escape bare `<` symbols

MDX parses < as the start of a JSX tag. In non-HTML/JSX contexts, < mustbe replaced with < .

Common triggering scenarios and fixes:

❌ 影响版本 <1.123.17 的用户    → 报错：Unexpected character `1`
✅ 影响版本 &lt;1.123.17 的用户

❌ 当 x < 10 时               → 报错：Unexpected character `1`
✅ 当 x &lt; 10 时

❌ Array<string>              → 报错：解析为 JSX
✅ `Array<string>`            → 用行内代码包裹
✅ Array&lt;string&gt;         → 或用 HTML 实体

Note: inside code blocks (```) the < does not need to be escaped; it is safe inside code blocks.

Rule 2: Code block language identifiers must be languages supported by Shiki

The language identifier of a code block (```language) must use a language name known to Shiki, otherwise the build will report ShikiError: Language "xxx" is not included in this bundle。

Safe language identifiers (commonly used):

text, plaintext, txt, bash, shell, sh, javascript, js, typescript, ts,
python, py, json, yaml, yml, html, css, sql, go, rust, java, c, cpp,
markdown, md, diff, toml, xml, graphql, docker, dockerfile, ini, csv

Nonexistent/unsafe language identifiers:

❌ ```td        → Shiki 无此语言，改用 ```text
❌ ```nushell   → Shiki 无此语言，改用 ```shell 或 ```text
❌ ```hcl       → 改用 ```text

Principle: If you are unsure whether a language is supported by Shiki, always use text。

Rule 3: Curly braces `{}` are parsed as JSX expressions

In MDX, { and } in the body text are evaluated as JSX expressions.

❌ 格式：{名称}    → 报错：名称 is not defined
✅ 格式：\{名称\}  → 转义
✅ 格式：`{名称}`  → 用行内代码包裹

Note: inside code blocks, {} does not need to be escaped.

Notes

Depth over breadth: Better to include 10 high-quality news items than 30 low-quality ones
Verification over speed: Ensure each news item is verified through multiple sources
Interpretation over retelling: Provide technical interpretation and developer action recommendations
Deduplication over repetition: Check content published in the past 7 days to avoid duplicate coverage
English search + Chinese output: Using English keywords when searching is more effective
Stay objective and neutral: Avoid subjective evaluations and emotional expressions