All notable changes to the trend-radar skill suite will be documented in this file. The format follows Keep a Changelog and the suite uses Semantic Versioning.
- Initial public release of the trend-radar skill suite.
trend-monitor— multi-platform hotspot sweep + auto-extracted emerging keywords across 11 sources (zhihu, weibo, douyin, bilibili-hot-search, baidu, toutiao, thepaper, ifeng, tieba, cls-hot, wallstreetcn-hot).trend-deep-dive— single-topic analysis withtrend/lifecycle/viral/predictmodes; includes decision tree for picking analysis type.trend-alert— anomaly detection + push to 9 channels (飞书 / 钉钉 / 企业微信 / Telegram / Slack / Bark / ntfy / Email / Generic Webhook) with channel-correct formatting and rate limiting.viral-forge— platform-native content generation for 小红书 / 抖音 / B站 / 微博 / 知乎 / 公众号; ships with 36 proven title patterns and 合规 guardrails.trend-report— daily / weekly / monthly / period-vs-period digest reports with three audience templates (执行摘要 / 媒体行业 / 投放团队).docs/mcp-tools.md— fullmcp__trendradar__*tool catalog..github/workflows/validate.yml— CI gate usingskills-ref validate.
- Compliant with the Agent Skills specification at https://agentskills.io/specification.
- Each skill is self-contained —
viral-forgecarries its ownplatform-playbook.md; cross-skill references are informational only.