Search and Monitor Twitter Trends by Keyword

Runs on-demand and scheduled keyword searches across Twitter without browser automation or official API, so you find high-engagement tweets and track topic shifts over weeks. It uses rnet to query Twitter's internal GraphQL search API directly, returning raw tweets that you can filter, aggregate, and analyze for actionable trends. Most teams either pay $5,000+/mo for Twitter API Premium or miss emerging conversations.

1. On-demand search: you specify a keyword or phrase, the skill searches Top or Latest mode and returns up to 200 tweets. 2. Filter and enrich: remove low-engagement results (< 10 likes), original tweets only, with mention of the keyword. 3. Generate report: list top tweets, extract dominant themes, and identify key authors. 4. Optional cron monitoring: define keywords in a config file; the skill runs scheduled searches, diffs against last run, and alerts on new high-engagement tweets. 5. Trend analysis: aggregate weeks of saved search data, group by week, and generate a narrative showing how sentiment and themes evolved. Output: structured search results, daily monitoring logs, and week-by-week trend narratives.

Requires rnet_twitter.py and valid twitter_cookies.json. SearchTimeline endpoint requires POST (GET returns 404); this is handled automatically. GraphQL query IDs rotate periodically; monitor for 403/404 errors and update ENDPOINTS dict in rnet_twitter.py when needed. Rate limit: ~300 requests/15 min; searching 200 tweets uses ~10 requests, safe for cron every 4 hours.