qmd by levineam/qmd-skill
npx skills add https://github.com/levineam/qmd-skill --skill qmd用于 Markdown 笔记、文档和知识库的本地搜索引擎。一次索引,快速搜索。
qmd search (BM25)。它通常是即时的,应作为默认选项。qmd vsearch(冷启动时可能非常慢)。qmd query。brew install sqlite (SQLite 扩展)$HOME/.bun/bin安装 Bun (macOS): brew install oven-sh/bun/bun
广告位招租
在这里展示您的产品或服务
触达数万 AI 开发者,精准高效
bun install -g https://github.com/tobi/qmd
qmd collection add /path/to/notes --name notes --mask "**/*.md"
qmd context add qmd://notes "Description of this collection" # 可选
qmd embed # 一次性执行以启用向量 + 混合搜索
**/*.md)。qmd search (默认): 快速关键词匹配 (BM25)qmd vsearch (最后手段): 语义相似性 (向量)。由于在向量查找之前需要本地 LLM 工作,通常较慢。qmd query (通常跳过): 混合搜索 + LLM 重排序。通常比 vsearch 更慢,并且可能超时。qmd search 通常是即时的。qmd vsearch 在某些机器上可能需要约 1 分钟,因为查询扩展可能每次运行都会将本地模型(例如 Qwen3-1.7B)加载到内存中;向量查找本身通常很快。qmd query 在 vsearch 的基础上增加了 LLM 重排序,因此对于交互式使用来说可能更慢且可靠性更低。qmd search "query" # 默认
qmd vsearch "query"
qmd query "query"
qmd search "query" -c notes # 搜索特定集合
qmd search "query" -n 10 # 更多结果
qmd search "query" --json # JSON 输出
qmd search "query" --all --files --min-score 0.3
-n <num>: 结果数量-c, --collection <name>: 限制在特定集合--all --min-score <num>: 返回高于阈值的所有匹配项--json / --files: 适合代理的输出格式--full: 返回完整文档内容qmd get "path/to/file.md" # 完整文档
qmd get "#docid" # 根据搜索结果中的 ID
qmd multi-get "journals/2025-05*.md"
qmd multi-get "doc1.md, doc2.md, #abc123" --json
qmd status # 索引健康状态
qmd update # 重新索引已更改的文件
qmd embed # 更新嵌入向量
自动化索引,以便在您添加/编辑笔记时结果保持最新。
qmd search),qmd update 通常就足够了(快速)。vsearch/query),您可能还需要 qmd embed,但这可能很慢。示例计划任务 (cron):
# 每小时增量更新(保持 BM25 新鲜):
0 * * * * export PATH="$HOME/.bun/bin:$PATH" && qmd update
# 可选:每晚刷新嵌入向量(可能很慢):
0 5 * * * export PATH="$HOME/.bun/bin:$PATH" && qmd embed
如果您的 Clawdbot/代理环境支持内置调度器,您可以在那里运行相同的命令,而不是使用系统 cron。
~/.cache/qmd/models/ (可通过 XDG_CACHE_HOME 覆盖)。qmd 搜索 您的本地文件(笔记/文档),这些文件是您明确索引到集合中的。memory_search 搜索 代理记忆(从先前交互中保存的事实/上下文)。memory_search 用于"我们之前决定/学习了什么?",qmd 用于"我的磁盘上的笔记/文档里有什么?"。每周安装次数
723
仓库
GitHub 星标数
655
首次出现
2026 年 1 月 28 日
安全审计
安装于
gemini-cli625
opencode625
codex622
openclaw619
github-copilot603
kimi-cli581
Local search engine for Markdown notes, docs, and knowledge bases. Index once, search fast.
qmd search (BM25). It's typically instant and should be the default.qmd vsearch only when keyword search fails and you need semantic similarity (can be very slow on a cold start).qmd query unless the user explicitly wants the highest quality hybrid results and can tolerate long runtimes/timeouts.brew install sqlite (SQLite extensions)$HOME/.bun/binInstall Bun (macOS): brew install oven-sh/bun/bun
bun install -g https://github.com/tobi/qmd
qmd collection add /path/to/notes --name notes --mask "**/*.md"
qmd context add qmd://notes "Description of this collection" # optional
qmd embed # one-time to enable vector + hybrid search
**/*.md).qmd search (default): fast keyword match (BM25)qmd vsearch (last resort): semantic similarity (vector). Often slow due to local LLM work before the vector lookup.qmd query (generally skip): hybrid search + LLM reranking. Often slower than vsearch and may timeout.qmd search is typically instant.qmd vsearch can be ~1 minute on some machines because query expansion may load a local model (e.g., Qwen3-1.7B) into memory per run; the vector lookup itself is usually fast.qmd query adds LLM reranking on top of vsearch, so it can be even slower and less reliable for interactive use.qmd search "query" # default
qmd vsearch "query"
qmd query "query"
qmd search "query" -c notes # Search specific collection
qmd search "query" -n 10 # More results
qmd search "query" --json # JSON output
qmd search "query" --all --files --min-score 0.3
-n <num>: number of results-c, --collection <name>: restrict to a collection--all --min-score <num>: return all matches above a threshold--json / --files: agent-friendly output formats--full: return full document contentqmd get "path/to/file.md" # Full document
qmd get "#docid" # By ID from search results
qmd multi-get "journals/2025-05*.md"
qmd multi-get "doc1.md, doc2.md, #abc123" --json
qmd status # Index health
qmd update # Re-index changed files
qmd embed # Update embeddings
Automate indexing so results stay current as you add/edit notes.
qmd search), qmd update is usually enough (fast).vsearch/query), you may also want qmd embed, but it can be slow.Example schedules (cron):
# Hourly incremental updates (keeps BM25 fresh):
0 * * * * export PATH="$HOME/.bun/bin:$PATH" && qmd update
# Optional: nightly embedding refresh (can be slow):
0 5 * * * export PATH="$HOME/.bun/bin:$PATH" && qmd embed
If your Clawdbot/agent environment supports a built-in scheduler, you can run the same commands there instead of system cron.
~/.cache/qmd/models/ (override with XDG_CACHE_HOME).qmd searches your local files (notes/docs) that you explicitly index into collections.memory_search searches agent memory (saved facts/context from prior interactions).memory_search for "what did we decide/learn before?", qmd for "what's in my notes/docs on disk?".Weekly Installs
723
Repository
GitHub Stars
655
First Seen
Jan 28, 2026
Security Audits
Gen Agent Trust HubWarnSocketWarnSnykPass
Installed on
gemini-cli625
opencode625
codex622
openclaw619
github-copilot603
kimi-cli581
头脑风暴技能:AI协作设计流程,将创意转化为完整规范与实施计划
77,000 周安装