Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →GitHub Trending 探索与分析。用于发现热门开源项目、技术趋势、开发者偏好，帮助理解技术社区的兴趣走向。使用场景包括“看看今天 GitHub 什么火了”、“Rust 最近热门项目”、“AI 领域趋势日报”、“这个方向有没有值得关注的开源项目”、“帮我做个 GitHub trending 分析”。
.claude/skills/majiayu000-github-trending/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-04 | ✗→✓ | ▲ Improved | -3% | 0% |
| case-13 | ✗→✓ | ▲ Improved | -2% | 0% |
| case-18 | ✗→✓ | ▲ Improved | 178% | 0% |
| case-19 | ✗→✓ | ▲ Improved | -4% | 0% |
| case-06 | ✓→✓ | ＝ Same ✓ | 258% | 0% |
2026 版核心认知：GitHub Trending 已是严重被 hype 和 AI slop 污染的榜单（Karpathy 等反复确认）。本 skill 的唯一价值是「结构化数据 + 严格过滤 + 写出真实 angle」，而不是再输出一份 list。
职责边界（严格遵守）：
/scout-to-article）/multi-ai-research）不适合：纯快速事实查询（直接让 Claude 回答即可）。
bash # 在 skill 目录执行；如果 skill 已安装到 ~/.claude/skills，也可以换成对应安装路径 cd /path/to/spellbook/skills/github-trending
# 今日总榜 TOP 8 python3 scripts/fetch_trending.py --since daily --limit 8
# Rust 周榜 python3 scripts/fetch_trending.py --since weekly --language Rust --limit 6
# 开发者榜 python3 scripts/fetch_trending.py --developers --since daily --limit 10 脚本输出结构化 JSON，绝不直接 WebFetch 原始 HTML。
reference/extended.md 中的「Hype 红旗」快速排除首次使用执行：
bashpython3 -m pip install -r requirements.txt
如果没有 requirements.txt，手动安装：
bashpython3 -m pip install requests beautifulsoup4 lxml
脚本位置：scripts/fetch_trending.py（相对本 skill 目录）。
支持参数见脚本 --help。所有错误都会以清晰 JSON 返回，绝不静默失败。
日报结尾必须包含「Angle」段，否则视为不合格输出。
好 Angle 特征：
reference/extended.md 里的评估框架写对比表。/loop 技能每天定时跑脚本 + 分析，输出到固定目录。/multi-ai-research。scout-to-article。所有详细框架（数据源、评估维度、hype 识别、趋势分类、好/坏 Angle 示例、scraper 维护提示、2026 社区共识）已移至 reference/extended.md。
日常触发时不需要读它。只有你要定制模板、维护脚本、或做高精度分析时再加载。
更新记录：2026-05 完整重构（引入可靠 scraper 层 + 严格 pipeline + 拆分 reference + 移除所有硬编码日期/旧趋势列表）。数据永远以脚本实时输出为准。
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 19,744 | 31,215 | +58% | 1 | 1 | 0% | 2,747 | 4,973 | +81% | 0 | 0 | — |
case-02 | fail→fail | 24,127 | 5,381 | -78% | 1 | 1 | 0% | 3,792 | 1,405 | -63% | 0 | 0 | — |
case-03 | fail→fail | 18,552 | 7,331 | -60% | 1 | 1 | 0% | 2,908 | 1,407 | -52% | 0 | 0 | — |
case-04 | fail→pass | 12,806 | 4,067 | -68% | 1 | 1 | 0% | 1,815 | 1,754 | -3% | 0 | 0 | — |
case-05 | fail→fail | 31,181 | 9,837 | -68% | 1 | 1 | 0% | 4,867 | 1,769 | -64% | 0 | 0 | — |
case-06 | pass→pass | 3,008 | 2,085 | -31% | 1 | 1 | 0% | 391 | 1,398 | +258% | 0 | 0 | — |
case-07 | fail→fail | 11,704 | 7,613 | -35% | 1 | 1 | 0% | 1,782 | 2,155 | +21% | 0 | 0 | — |
case-08 | fail→fail | 13,222 | 2,668 | -80% | 1 | 1 | 0% | 2,257 | 1,440 | -36% | 0 | 0 | — |
case-09 | fail→fail | 13,868 | 3,062 | -78% | 1 | 1 | 0% | 2,606 | 1,492 | -43% | 0 | 0 | — |
case-10 | pass→pass | 11,254 | 4,675 | -58% | 1 | 1 | 0% | 1,750 | 1,792 | +2% | 0 | 0 | — |
case-11 | pass→pass | 10,427 | 6,463 | -38% | 1 | 1 | 0% | 1,540 | 2,031 | +32% | 0 | 0 | — |
case-12 | pass→pass | 8,693 | 5,353 | -38% | 1 | 1 | 0% | 1,401 | 1,914 | +37% | 0 | 0 | — |
case-13 | fail→pass | 10,832 | 3,497 | -68% | 1 | 1 | 0% | 1,698 | 1,666 | -2% | 0 | 0 | — |
case-14 | fail→fail | 13,274 | 10,440 | -21% | 1 | 1 | 0% | 2,072 | 2,654 | +28% | 0 | 0 | — |
case-15 | fail→fail | 9,937 | 3,643 | -63% | 1 | 1 | 0% | 1,413 | 1,581 | +12% | 0 | 0 | — |
case-16 | pass→pass | 12,251 | 7,384 | -40% | 1 | 1 | 0% | 1,764 | 2,176 | +23% | 0 | 0 | — |
case-17 | fail→fail | 17,931 | 16,005 | -11% | 1 | 1 | 0% | 2,921 | 3,644 | +25% | 0 | 0 | — |
case-18 | fail→pass | 27,448 | 11,375 | -59% | 1 | 1 | 0% | 1,058 | 2,940 | +178% | 0 | 0 | — |
case-19 | fail→pass | 10,321 | 3,012 | -71% | 1 | 1 | 0% | 1,564 | 1,509 | -4% | 0 | 0 | — |
case-20 | pass→pass | 8,892 | 2,201 | -75% | 1 | 1 | 0% | 1,458 | 1,434 | -2% | 0 | 0 | — |
case-21 | pass→pass | 10,841 | 2,338 | -78% | 1 | 1 | 0% | 1,824 | 1,373 | -25% | 0 | 0 | — |
case-22 | fail→fail | 9,998 | 2,401 | -76% | 1 | 1 | 0% | 1,342 | 1,411 | +5% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted, and 18 counted toward the lift figure. The other 4 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +18 percentage points is the difference between those two pass rates over the 18 comparable cases. 3 cases got worse with the skill loaded, and they are included in that figure.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.