News For AI#

Fetches AI news, daily digests, and searches AI-related content. Outputs structured JSON (clean Markdown text + separated images/videos/links), ready for article writing or analysis.

Setup#

bash

pip install -r ${SKILL_DIR}/requirements.txt

Command Selection#

Pick the right command based on user intent:

User intentCommandExample最新AI新闻/资讯/热点/行业动态news``news --limit 10AI日报/每日汇总/今日速递daily``daily --limit 3查找特定话题/产品/模型/技术search``search "GPT" --type news新闻+日报一起获取all``all --limit 10

CLI Reference#

bash

python ${SKILL_DIR}/news_cli.py <command> [options]

Commands and Options#

bash

# AI新闻资讯
python ${SKILL_DIR}/news_cli.py news [--limit N] [--no-content]

# AI日报
python ${SKILL_DIR}/news_cli.py daily [--limit N] [--no-content]

# 搜索（关键词必填）
python ${SKILL_DIR}/news_cli.py search "关键词" [--type TYPE] [--limit N] [--no-content]
#   --type: news / products / models / mcp / all (default: all)

# 新闻+日报
python ${SKILL_DIR}/news_cli.py all [--limit N] [--no-content]

OptionEffect--limit N返回条数，默认 20--no-content跳过正文抓取，仅返回标题+摘要（快 20x，适合浏览选题）--type TYPE搜索范围过滤，仅 search 命令可用

Article Writing Workflow#

For gathering source material to write AI articles:

1. 选题 — 快速扫标题： news --limit 20 --no-content
2. 深挖 — 获取选中文章的完整正文+图片+链接： news --limit 5
3. 定向搜索 — 补充特定话题素材： search "Sora" --type news --limit 3

Output Format#

All commands output JSON to stdout, errors to stderr (exit code 1).

news / daily#

json

{
  "type": "news",
  "timestamp": "2026-03-30 15:00:00 +0800",
  "count": 5,
  "data": [
    {
      "id": 26659,
      "title": "...",
      "description": "摘要...",
      "url": "https://...",
      "create_time": "2026-03-30 14:52:05",
      "pv": 5426,
      "content": "正文纯文本（Markdown格式）...",
      "images": [{"src": "https://...", "alt": "..."}],
      "videos": [{"src": "https://..."}],
      "links": [{"text": "...", "href": "https://..."}]
    }
  ]
}

--no-content 模式下 content、 images、 videos、 links 字段不会出现。

search#

json

{
  "type": "search",
  "keyword": "千问",
  "search_type": "all",
  "news": {"type": "news", "count": 5, "data": [...]},
  "products": {"type": "products", "count": 3, "data": [...]},
  "models": {"type": "models", "count": 10, "data": [...]},
  "mcp": {"type": "mcp", "count": 1, "data": [...]}
}

For complete field definitions of each item type, see schemas.md.