メニュー
Search the Obsidian vault for all notes related to a topic. First step in the brain brief pipeline.
デフォルトでは、最初にソースを確認する Prompt が選択されています。直接コマンドに切り替えるか、ローカルコピーをダウンロードすることもできます。
インストールを決める前に、SKILL.md と SkillsMP に表示されている付属ファイルをお読みください。
SOC 職業分類に基づく
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。
直接コマンドでは確認用 Prompt が省略されます。実行前にソースを確認してください。
npx skills add https://github.com/earlyaidopters/gumroad-resources --skill brain-searchコマンドは1行のまま表示されます。コピー前に横へスクロールして全体を確認してください。
ローカルで確認しますか？SkillsMP が現在取得できるファイルをダウンロードできます。
SKILL.md を表示中
Research and diagnose which AI personal-assistant operating model fits a user, including whether to stay put, use Grok Bot, Hermes Agent, Claude Cowork, ChatGPT Work, OpenClaw, or build an owned system. Use for assistant comparisons, switch decisions, product deep dives, and DIY assistant architecture planning.
Personal Higgsfield replacement. Routes a plain-English idea to the best pay-as-you-go image or video model, refines the prompt for that model, calls the API via scripts/generate.py, saves the output, and logs cost to a running receipts ledger. Use when the user says "generate an image", "make a video of", "generate", or describes any visual they want created.
Build premium editorial DOM websites — expressive serif display type, soft warm canvases, rounded media plates, alternating feature rows, card rails, calm scroll reveals. No WebGL, no scroll-scrubbed video. Use for product/SaaS landing pages, brand sites, hospitality/outdoor brands, or any request for a "clean", "editorial", "premium", or "magazine-like" site. For 3D/immersive/scroll-scrubbed sites use the immersive-web skill instead.
| name | brain-search |
| description | Search the Obsidian vault for all notes related to a topic. First step in the brain brief pipeline. |
Search the Obsidian vault for everything related to: "$ARGUMENTS"
Update this path to point to YOUR Obsidian vault: VAULT_PATH = "~/path/to/your/obsidian/vault"
Write results to outputs/brain/!date +%Y-%m-%d/{slug}/01_search_results.md where {slug} is the topic lowercased with hyphens.
Format:
# Brain Search: [Topic]
*Searched !`date +%Y-%m-%d`*
*Vault: [Your Vault Name]*
## Notes Found
### 1. [Note Title]
**Path:** [folder/filename.md]
**Tags:** [any tags]
**Preview:** [First 2-3 lines of content]
### 2. [Note Title]
...
## Summary
- **Total notes found:** [X]
- **Areas covered:** [which vault folders had matches]
- **Key tags:** [relevant tags found across matches]
Search broadly. Include partial matches and notes that reference the topic indirectly through wikilinks.