AITOP AI News Assistant

Problem

AI updates are scattered across product launches, model releases, papers, and industry funding news. When an assistant scrapes raw pages directly, results can mix summaries, miss source links, and have inconsistent fields. This skill wraps AITOP’s public API into a structured news query workflow for engineers and agents that need reliable AI industry updates.

How It Works

• Routing: choose GET /items, GET /daily, or GET /items/{id} based on intent. Use curated mode by default and only switch to mode=all when the user explicitly asks for all results.
• Filtering: prefer server-side parameters such as since, q, and category instead of fetching everything and filtering locally. Supported categories include ai-models, ai-products, industry, paper, and tip.
• Display: use detailUrl as the default reading link and sourceUrl for citation, verification, or “view source” requests. Do not treat AI summaries as source text.
• Details: when expanding an item, read fields such as recommendReason, bodyOriginal, and duplicates. If any *Truncated=true field appears, append readMore.url and state that the content is an excerpt.

Boundaries

This skill is suitable for querying AITOP-curated AI news, daily briefs, and item details, not for arbitrary web crawling or generating new analysis. If the API fails, say AITOP is temporarily unavailable rather than inventing cached results. Output any notice once, verbatim, at the end of the response, and keep subscription keys out of user-facing output or logs.