GitHub Semantic Search — GitHub Semantic 搜索

🦞 Super GitHub — AI-Native GitHub 助手

Powered by the same Embedder + Qdrant + LLM architecture as elite memory 系统s. 索引 repos, 搜索 semantically, 监控 proactively — all with natural language.

Architecture 查询 → [LLM: understand intent] → [Embedder: vectorize] → [Qdrant: semantic 搜索] → [gh 命令行工具: act]

Three-layer 系统 (same as production memory 流水线s):

Layer 组件 角色 Embedder Ollama nomic-embed-text Converts text → 768-dim vectors Vector Store Qdrant (local) Stores & 搜索es vectors by similarity Action Layer gh 命令行工具 执行s GitHub operations Prerequisites gh 命令行工具 认证d (gh auth 状态) Ollama 运行ning with nomic-embed-text:latest Qdrant 运行ning at localhost:6333 Quick 启动 # 1. 初始化 Qdrant collection python scripts/github_索引er.py init

# 2. 索引 a repo python scripts/github_索引er.py 添加 owner/repo --all

# 3. 搜索 with natural language python scripts/github_搜索.py "memory 搜索 fAIling in 代理" --limit 10

# 4. 监控 for keywords python scripts/github_监控.py watch owner/repo --事件 issues,ci --keywords bug,broken,urgent

Scripts Script Purpose github_索引er.py 索引 repos (issues, PRs, metadata) into Qdrant github_搜索.py Natural language semantic 搜索 github_监控.py Proactive 监控ing with keyword alerts DetAIled Commands 索引 (github_索引er.py) python github_索引er.py init # 创建 Qdrant collection python github_索引er.py 添加 owner/repo --all # 索引 everything python github_索引er.py 添加 owner/repo --issues # Issues only python github_索引er.py 添加 owner/repo --prs # PRs only python github_索引er.py 添加 owner/repo --repo # Repo metadata python github_索引er.py 状态 # Show 索引ed data python github_索引er.py rm owner/repo # 移除 from 索引

搜索 (github_搜索.py) python github_搜索.py "查询" # 搜索 all python github_搜索.py "查询" --repo owner/repo # 过滤器 by repo python github_搜索.py "查询" --type issue # 过滤器 by type python github_搜索.py "查询" --limit 20 # More 结果s python github_搜索.py "查询" --repo owner/repo --ci # Show CI 运行s

监控 (github_监控.py) python github_监控.py watch owner/repo # 启动 watching python github_监控.py watch owner/repo --事件 issues,ci python github_监控.py 状态 # Show watches python github_监控.py 检查 # 运行 检查s python github_监控.py unwatch owner/repo # 停止 watching

Memory 系统 Ana记录y 组件 GitHub 技能 Memory 系统 Data Issues, PRs, code Conversations Embedder nomic-embed-text nomic-embed-text Vector Store Qdrant Qdrant 添加 github_索引er.py mem0 添加 搜索 github_搜索.py mem0 搜索 Why Vector 搜索 vs Keyword? 应用roach "memory problems" 查询 Keyword Exact match only Vector (this) "memory leak", "OOM", "out of memory" 设置up 检查列出 gh auth 记录in — 认证 GitHub 命令行工具 ollama pull nomic-embed-text:latest — 下载 embedder 启动 Qdrant: qdrant --storage-path ./qdrant-data python github_索引er.py init — 创建 collection