competitor-news-monitor

Define a watchlist of companies with domains, aliases, and event categories, then build source coverage from official newsrooms, changelogs, and filings. The Skill stores this as a watch contract and creates a cron job that runs incremental checks and delivers digests of material events.

Run the one-time foreground setup to freeze the watchlist and materiality threshold, then the Skill creates a cronjob with your chosen schedule and delivery destination. Each tick collects news since the last cutoff, deduplicates events, and sends a cited digest.

Coverage includes official newsrooms, blogs, changelogs, pricing and product pages, regulatory filings, investor relations, status and security pages, reputable trade and financial press, and job postings as weak supporting evidence. Gaps in coverage are documented explicitly.

No. It deduplicates syndicated stories, rewrites, and URL variants into single underlying events, then scores each against a materiality threshold. When no material events occur, it stays silent unless you requested a periodic all-clear.

A source failure is recorded as unknown coverage, never treated as "no news." The collection cutoff only advances for successfully covered sources, so failed sources are rechecked on the next run and reported as coverage gaps.

Avoid it for one-off company research, where direct web search and extraction are faster, and for plain feed reading, which a feed watcher handles better. It is designed for recurring, threshold-based monitoring of a fixed company set.