メニュー
qmd - Search local markdown notes, documentation, and Claude session history. Use when user says "qmd", "search notes", "find docs", "search sessions", "local search", "knowledge base search", or needs to retrieve markdown files.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
Generate clean, minimal technical SVG diagrams in a consistent Cloudflare-inspired style. Use when creating architecture diagrams, flow diagrams, or component diagrams for blog posts and documentation.
Guides creation and improvement of best-practice agent skills following the open format specification. Covers frontmatter, directory structure, progressive disclosure, reference files, rules folders, degrees of freedom, content patterns, executable scripts, MCP tool references, evaluations, cross-model testing, and a ten-dimension audit protocol for existing skills. Use when creating a new skill, authoring SKILL.md, setting up a rules-based audit skill, structuring a skill bundle, writing scripts inside a skill, evaluating a skill, improving or rewriting an existing skill, or asking "how to write a skill", "improve this skill", "audit my skill", or "review this SKILL.md".
Diagnoses Erivault application behavior across product docs, repo code, Cloudflare, Neon Postgres, queues, Durable Objects, R2 assets, AI Gateway, media workflows, browser, and mobile capture state. Use when investigating Erivault bugs, asking "what happened", checking logs, verifying deploys, debugging capture evidence, or designing testable diagnostics.
Design a new product module's full screen set in the Erivault `erivault-web` Paper file, reusing the established design system (tokens, shell, components) and grounded in the module's ADR + implementation plan. Covers researching the existing system, extracting authoritative tokens, planning the state×role×platform matrix, building screens via duplicate-and-adapt, and tying into existing surfaces. Use when designing a new module or surface (tenancies, inspections, renewals, capture, packs...) in Paper, extending the erivault-web design file, building all states/roles for a feature on web and mobile, or when the user says "design the X module", "design a new flow", "design X in the Paper board", "add the screens for X", or "make it consistent with our design system".
View and edit cmux settings in ~/.config/cmux/cmux.json. Use when the user wants to change cmux preferences (appearance, sidebar, notifications, automation, browser, shortcuts), set a value by JSON path, validate the file, open it in an editor, or look up which keys cmux recognizes. Triggers on '/cmux-settings', 'change cmux setting', 'set <something> in cmux', 'cmux config', 'cmux.json', or 'rebind a cmux shortcut'.
Perform exhaustive code reviews using multi-agent analysis and dynamic skill discovery
| name | qmd-search |
| description | qmd - Search local markdown notes, documentation, and Claude session history. Use when user says "qmd", "search notes", "find docs", "search sessions", "local search", "knowledge base search", or needs to retrieve markdown files. |
Fast local search engine for markdown notes, documentation, and Claude session history.
# Keyword search (instant, preferred)
qmd search "authentication flow"
# Get full document
qmd get "path/to/file.md"
# Search specific collection
qmd search "react hooks" -c claude-sessions
qmd search (Default - Use This)BM25 keyword matching. Instant results.
qmd search "error handling"
qmd search "typescript generics" -n 10
qmd search "api design" --json
qmd vsearch (Use Sparingly)Vector similarity search. Slower - loads local models on each run.
qmd vsearch "concepts similar to dependency injection"
qmd query (Avoid)Hybrid search with LLM reranking. Slowest, often less reliable.
qmd search "query" # default keyword search
qmd search "query" -n 10 # more results
qmd search "query" -c collection # specific collection
qmd search "query" --json # JSON output
qmd search "query" --files # file paths only
qmd search "query" --full # full document content
qmd search "query" --all --min-score 0.3 # all matches above threshold
qmd get "path/to/file.md" # full document
qmd get "path/to/file.md:50" # from line 50
qmd get "path/to/file.md" -l 100 # max 100 lines
qmd multi-get "journals/2025-01*.md" # glob pattern
qmd multi-get "doc1.md, doc2.md" # multiple files
qmd collection list # show all collections
qmd collection add /path --name notes --mask "**/*.md"
qmd ls # list files in collections
qmd ls collection-name # list files in specific collection
qmd status # index health
qmd update # re-index changed files
qmd embed # update vector embeddings
| Flag | Description |
|---|---|
--json | JSON with snippets |
--files | docid, score, filepath, context |
--csv | CSV output |
--md | Markdown output |
--xml | XML output |
--full | Complete document content |
| Option | Description |
|---|---|
-n <num> | Number of results (default: 5) |
-c, --collection <name> | Filter to collection |
--all | Return all matches |
--min-score <num> | Minimum similarity score |
--line-numbers | Add line numbers |
qmd search - instant BM25 keyword matchingqmd vsearch sparingly - slow due to model loadingqmd query - hybrid search often unreliable--json or --files for agent-friendly output-c collection to narrow scope when you know the targetIndex location: ~/.cache/qmd/index.sqlite