ai-news-collectors

**Judgement:** A well-designed, task-focused skill that should perform strongly at discovering and ranking current AI news for Chinese-speaking users, especially when web tools are available and responsive. **Strengths & Value:** - Uses a detailed, multi-dimensional search strategy (newsletters, community sites, funding, research, policy) instead of naive single-query search, likely improving coverage and catching community-viral items. - Includes explicit heuristics for heat scoring, de-duplication, and cross-validation across sources, which should make the rankings more meaningful and reduce redundancy. - Produces structured, consistent Chinese summaries with links and brief metadata (counts, dimensions covered), which is convenient for dashboards, reposting, or manual review. **Key Risks / Limitations:** - Heavy reliance on multiple web_search and web_fetch calls (8–12+ per run) may increase latency, cost, and susceptibility to transient network or API errors. - Quality and completeness depend heavily on external search results and the maintained source patterns; if search quality degrades or sources change, coverage may suffer. - No explicit safety or misinformation filtering layer beyond using multiple sources; controversial or low-quality sources could appear if surfaced by search. **Recommended Scenarios:** - Building a routine "AI this week" or "today’s AI highlights" feature for Chinese-speaking users, Slack/Discord groups, or internal newsletters. - Assisting analysts, PMs, or founders who need quick situational awareness of current AI products, funding, and research, without manually scanning many sites. - Tracking community buzz around open-source AI tools, agents, and meme events that are often missed by traditional news aggregators. **Less Ideal For:** - Use cases requiring guaranteed low latency or strict resource limits (due to many web calls). - Applications needing rigorous fact-checking, bias analysis, or strong content filtering beyond straightforward aggregation and ranking.