ClawHub Daily Skill Insights

Problem It Solves

AI Agent Skill ecosystems update quickly, but manually scanning ClawHub can be noisy, repetitive, and hard to prioritize. When the signal is buried, the Skills worth testing may be missed too. skill-daily is for engineers, team leads, and creators who need a regular digest of useful Skills.

How It Works

It fetches ClawHub Top 200 Skills via the Convex API and computes signals such as stars, downloads, installsCurrent, comments, and capabilityTags. Instead of serving one fixed ranking, it rotates among four views—trending, quality, newcomers, and panorama—using date % 4. A rolling 10-day snapshot deduplicates previous recommendations, and 7 scenario tags such as office automation, dev tools, and content creation help rank relevant results. The briefing uses a one-line Chinese explanation with the English original collapsed for quick scanning.

Each digest typically includes:
- 8–10 selected Skills with metrics and recommendation rationale
- Top 3 detailed notes for quick evaluation
- Feishu card and document kept to 200–400 characters for fast decision-making

Boundaries

It focuses on recommendation analysis only; it does not download, install, or deploy Skills. The data source is limited to the ClawHub Convex API. For SkillHub-specific analysis, immediate single queries, or deep Chinese-Skill reviews, choose another tool. Raw snapshots and custom metrics can also be read directly from the JSON output. Pushing requires configured credentials or manual triggering, and scheduled use depends on an external task runner.