360 Secure SkillHub
360 安全技能中心

AI Product Radar

一条命令运行 AI 产品雷达：抓取内置 7 个技术 RSS 源，去重并按关键词筛选 AI 相关内容，提取产品名称与类别、对发布信号打分，用 Playwright 截图（不可用时生成信息卡），最终输出排名趋势报告、结构化 JSON、截图与原始条目。

terrycarter1985
terrycarter1985
热度 13.44
最近更新 2026-08-05 22:02:22
分类：数据依赖：代码执行 / 联网浏览搜索 / 读写文件 / 第三方 API应用场景：市场分析 / 报告撰写 / 研究 / 数据处理 ...

安装技能

复制下面的话术发给你的 Agent，它会通过 secure-skills 检索并安装该技能。

执行命令安装此 skill： npx secure-skills-cli add ai-product-radar
Metadata
name
ai-product-radar
description
One-click AI product launch monitoring pipeline. Use when tracking new AI product releases, monitoring competitor launches, generating trend reports from tech RSS feeds, or running automated product discovery. Integrates RSS monitoring, product info enrichment, screenshot capture, and trend analysis into a single command. Triggers on "product radar", "AI launch monitor", "track new AI products", "product release tracking", "AI trend report", or similar product intelligence workflows.

AI Product Radar

Automated pipeline that monitors AI product launches across tech RSS feeds, enriches product data, captures screenshots, and produces a ranked trend report.

Quick Start

python3 scripts/ai_product_radar.py --output ./radar-output

Output:

  • report.md — human-readable trend report with ranked products, categories, themes
  • products.json — full structured data (all products, scores, categories)
  • screenshots/ — product page screenshots (or info cards if no browser)
  • raw/rss_items.json — raw feed items

Options

Flag Default Description
--output, -o ./ai-radar-output Output directory
--feeds, -f built-in list JSON file with custom RSS URLs
--limit, -l 15 Max screenshots to capture
--days, -d 3 RSS lookback window
--no-screenshots off Skip screenshot stage
--query, -q AI auto-detect Filter term (narrows results)

Pipeline Stages

  1. RSS Monitoring — Fetches 7 built-in feeds (Product Hunt, TechCrunch AI, The Verge AI, VentureBeat AI, HN Show, MIT Tech Review, Ars Technica), deduplicates, and filters for AI relevance via keyword matching.
  2. Product Enrichment — Extracts product name, detects categories (LLM, image gen, dev tools, agents, enterprise, etc.), scores launch signals.
  3. Screenshot Capture — Captures product pages with Playwright if available; falls back to generated info cards.
  4. Trend Analysis — Ranks products by signal score + recency, aggregates category/source distributions, extracts key themes.

Custom Feeds

Create a JSON file:

["https://example.com/feed.xml", "https://another.com/rss"]

Then pass --feeds my-feeds.json.

Scheduling

Set up a cron job for daily monitoring:

# Every day at 9:00 AM
python3 /path/to/scripts/ai_product_radar.py -o ~/radar/$(date +%Y-%m-%d)

Extending

  • Add RSS feeds to DEFAULT_FEEDS in the script or use --feeds.
  • Adjust AI_KEYWORDS for broader/narrower filtering.
  • Add categories in category_patterns inside search_product_info().
  • The script uses only Python stdlib + optional Playwright; no API keys required.