SkillHub Daily Domestic Insights

Problem

SkillHub.cn hosts a large pool of skills, but engineers usually need more than a hot list: they need to know which skills fit domestic workflows, active projects, and recent tasks. Manually checking rankings, categories, and search results is noisy and can easily miss new or actively updated skills. SkillHub Daily targets this filtering cost and missing context.

How It Works

It scans daily using skillhub CLI ranking and search APIs, merging hot, trending, featured, category long tail, and keyword supplements into a deduplicated candidate set of about 550 items. It then ranks candidates by signals such as trend surge, new releases, scenario fit, memory collision, domestic fit, active developers, and official verification. Memory collision reads local project_memory.md, topics.md, and user_profile.md, then weights matches against skill names, descriptions, and categories to align recommendations with current work.

The result is a daily snapshot, recommended JSON, and a structured Chinese brief, delivered separately to Obsidian, IMA FIM, and Feishu. Each destination is isolated, so one failure does not block the others.

Boundaries

It is better suited for daily domestic skill discovery and knowledge capture, not installation, security auditing, or dependency management. Its output depends on skillhub auth, Python, local memory files, and configured sync targets; missing setup can reduce recommendation accuracy or delivery coverage.