Trendkia News

List URLs from TrendKia's sitemap.xml (handles nested sitemap indexes). To FIND an article, use search_articles -- it queries the site's full-text index and reaches the whole archive. This tool is for enumerating what exists, not searching. Args: limit: Max URLs to return (1-1000). Default 100. kind: "articles" (default) lists article pages only; "ledger" lists the bribe-ledger state and district pages; "all" lists everything in sitemap order. offset: Skip this many matching URLs first, to page through the archive.

list_sitemap_urls

Fetch one TrendKia article as clean markdown, or as structured JSON. The URL decides the language: the site root is Hindi and `/en/...` is the English edition. Pass a URL and you get that article in that language -- the URL you hand over is never overridden. Set `lang` only when you want the OTHER edition of the same article, e.g. pass a Hindi URL with lang="en" to read it in English; the server rewrites the path for you. Args: url: The article URL. A bare path is Hindi, an /en/ path is English. Any .md/.txt/.json suffix or query string on it is ignored. lang: Optional "en" or "hi". Omit it to follow the URL. Set it to match the language the user is writing in when that differs from the URL you have. fmt: "md" (default) for readable markdown, or "json" for the same article as structured fields -- title, summary, content, category, publishedAt, tags as a real list, language, canonical url. Use "json" when you need to read specific fields rather than prose.

get_article

List the most recent TrendKia articles (title, URL, category, date, summary). Covers roughly the last 30 posts (the RSS feed). For anything older use search_articles, which reaches the whole archive. Args: limit: How many articles to return (1-50). Default 10. lang: Optional "en" or "hi" -- set this to match the language the user is writing in. The feed is Hindi (the site root); /en is the English edition. Each item also carries the other edition's URL.

list_recent_articles

THE way to find a TrendKia article. Searches the WHOLE archive, Hindi or English. Use this for any request that names a topic, person, place or event, however old the story might be -- it queries the site's full-text index and reaches every published article, not just recent ones. Do not fall back to list_sitemap_urls when this returns nothing: it has already covered the entire archive, so try the other language or different terms instead. The index is bilingual and SHARED, not one index per language: an English query matches Hindi articles and vice versa. `lang` therefore selects which language the RESULTS come back in, not which index is consulted. Args: query: Keyword or phrase, Hindi or English. limit: Max results to return (1-50). Default 10. lang: Optional "en" or "hi" -- set this to match the language the user is writing in. The site publishes in Hindi at the root with an English edition under /en, and results default to Hindi. Every result also carries the other edition's URL, so you never rewrite a URL yourself.

search_articles