Search Layer

This is a serious attempt at fixing search quality by throwing four engines at the problem: Brave, Exa, Tavily, and Grok. It classifies your intent first (factual, news, tutorial, comparison, etc.), then generates sub-queries, runs them in parallel across sources, deduplicates by URL, and scores results based on keyword match, freshness, and domain authority. The weights shift depending on what you're asking for. It can pull GitHub issue threads when it finds them in status queries, which is handy for tracking bugs or features. The synthesis layer tries to cluster results by theme instead of just dumping a list. It's meant to replace raw web_search calls entirely and handles both English and Chinese technical queries with synonym expansion.

skills.sh installs: 145
GitHub stars: 436
last updated: Aug 21, 2026

View on GitHub

Install

$ npx -y skills add blessonism/openclaw-search-skills --skill search-layer --agent claude-code

Installs into .claude/skills of the current project.

Files

Search Layer v2.2 — 意图感知多源检索协议

四源同级：Brave (web_search) + Exa + Tavily + Grok。按意图自动选策略、调权重、做合成。

执行流程

用户查询
    ↓
[Phase 1] 意图分类 → 确定搜索策略
    ↓
[Phase 2] 查询分解 & 扩展 → 生成子查询
    ↓
[Phase 3] 多源并行检索 → Brave + search.py (Exa + Tavily + Grok)
    ↓
[Phase 4] 结果合并 & 排序 → 去重 + 意图加权评分
    ↓
[Phase 5] 知识合成 → 结构化输出

Phase 1: 意图分类

收到搜索请求后，先判断意图类型，再决定搜索策略。不要问用户用哪种模式。

意图	识别信号	Mode	Freshness	权重偏向
Factual	"什么是 X"、"X 的定义"、"What is X"	answer	—	权威 0.5
Status	"X 最新进展"、"X 现状"、"latest X"	deep	pw/pm	新鲜度 0.5
Comparison	"X vs Y"、"X 和 Y 区别"	deep	py	关键词 0.4 + 权威 0.4
Tutorial	"怎么做 X"、"X 教程"、"how to X"	answer	py	权威 0.5
Exploratory	"深入了解 X"、"X 生态"、"about X"	deep	—	权威 0.5
News	"X 新闻"、"本周 X"、"X this week"	deep	pd/pw	新鲜度 0.6
Resource	"X 官网"、"X GitHub"、"X 文档"	fast	—	关键词 0.5