OpenSEO MCP
OFFICIAL SERVERSEO research for an assistant: keywords, SERPs, rankings, backlinks, local listings and site audits.
Lets an assistant search, analyse and summarise trending news and RSS data collected by TrendRadar
By sansan (individual) Python Data and analyticsWeb search
68 popularity score ★ 64k
TrendRadar, by sansan0, is a self-hosted trend and public-opinion monitor: it collects trending lists from many (mostly Chinese) platforms plus RSS feeds, filters them by keyword and pushes reports to channels such as WeChat Work, Feishu, DingTalk, Telegram, email, ntfy, Bark and Slack. Its MCP server lets an assistant work with the news data TrendRadar has collected.
The tools return the latest crawled news, trending topics and RSS items, search news and RSS by keyword, fetch news for a date range, and analyse topic trends, sentiment and related stories, compare periods, merge similar stories across platforms and generate daily or weekly summaries. Other tools read the full text of an article from a URL, trigger a crawl, sync data from remote storage, show configuration and status, and send a message to configured notification channels.
The MCP server needs local news data: you run TrendRadar yourself (the project ships test data to try it). Run it from the cloned project with uv in stdio mode, or start the HTTP service (start-http.sh, http://localhost:3333/mcp). A separate Docker image, wantcat/trendradar-mcp, runs the MCP service alongside the wantcat/trendradar crawler image.
resolve_date_rangeget_latest_newsget_trending_topicsget_latest_rsssearch_rssget_rss_feeds_statusget_news_by_dateanalyze_topic_trendanalyze_data_insightsanalyze_sentimentfind_related_newsgenerate_summary_reportaggregate_newscompare_periodssearch_newsget_current_configget_system_statuscheck_versiontrigger_crawlsync_from_remoteget_storage_statuslist_available_datesread_articleread_articles_batchget_channel_format_guideget_notification_channelssend_notification