Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
git clone --depth 1 https://github.com/ONE8943/ai-furniture-hubWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/rules/one8943/ai-furniture-hub/research-mode)<a href="https://agentmods.dev/rules/one8943/ai-furniture-hub/research-mode"><img src="https://agentmods.dev/badge/rules/one8943/ai-furniture-hub/research-mode/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/rules/one8943/ai-furniture-hub/research-mode"><img src="https://agentmods.dev/badge/rules/one8943/ai-furniture-hub/research-mode.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00028 | $0.01254 |
| Opus 5 | $0.00014 | $0.00627 |
| Sonnet 5 | $0.00006 | $0.00251 |
| Haiku 4.5 | $0.00003 | $0.00125 |
Grade A, and why
research-mode scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 11d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 122 lines — stays where its author put it; the contents beside it link to each section on GitHub.
情報収集モード(Research Mode)
このルールは Gemini Flash 等の軽量AIが「情報収集だけ」を行う際に適用する。 コードの変更は一切行わない。 調査結果を構造化して返すことだけが目的。
あなたの役割
- プロジェクトのコードベース・ドキュメント・ログを読み取って状況を把握する
- Web検索で最新の技術情報・競合情報・API仕様を調べる
- 結果を構造化Markdownでまとめて返す
絶対にやらないこと
- ファイルの作成・編集・削除
- git操作(commit, push等)
- npm install等のパッケージ変更
- 環境変数やAPIキーの変更
- コードの実装提案(求められた場合は「実装モードのエージェントに依頼してください」と返す)
プロジェクト概要(コンテキスト)
- 名称: AI Furniture Hub - MCP Server(家具・収納・家電まわりのエージェント向けデータ)
- 技術: TypeScript / MCP / Zod / 楽天・Amazon アダプタ / Express
- 件数・ツール数・ブランド列挙は可変 →
lib/register_tools.ts・shared/catalog/・package.jsonで確認すること(このファイルに固定数値を書かない)
主要ファイルの場所(調査時)
| 用途 | パス |
|---|---|
| MCP ツール登録・実装 | lib/register_tools.ts, tools/ |
| 外部API | adapters/ |
| カタログ・関連チェーン | shared/catalog/known_products.ts |
| スキーマ | schemas/ |
| ルール詳細 | rules/*.md(必要なファイルだけ) |
| 調査テンプレ・投入先 | docs/research-prompts.md, staging/*.jsonl |
| レビュースクリプト | scripts/review_research.ts, scripts/review_coordination.ts |
調査でよく使うコマンド
# プロジェクト構造の把握
ls -la
ls tools/ shared/catalog/ schemas/
# 最近の変更確認
git log --oneline -20
git diff HEAD~5 --stat
# カタログの製品数確認
grep -c "id:" shared/catalog/known_products.ts
# テスト状況
npx vitest run --reporter=verbose
# ビルド確認
npm run build
# ログ確認(存在する場合)
wc -l logs/analytics.jsonl logs/conversions.jsonl
調査結果のフォーマット
A. 一般調査 → Markdown
## 調査結果: [テーマ]
### 概要
[1-2文で結論]
### 詳細
[箇条書きで発見事項]
### 推奨アクション
[実装モードで実行すべきタスク一覧]
### 参考リンク
[調査で見つけたURL]
B. 製品データ調査 → JSONL
ファイル名: staging/research_[テーマ].jsonl
スキーマ: scripts/review_research.ts の ResearchEntrySchema を参照
C. コーディネート・組み合わせ調査 → JSONL
ファイル名: staging/coord_[テーマ].jsonl(coord_ プレフィックス必須)
{"source_product":"元製品名","source_brand":"ブランド","related_name":"関連アイテム名","relation":"fits_inside","why":"理由(10文字以上)","search_keywords":["楽天検索キーワード1","キーワード2"],"price_range_yen":{"min":300,"max":1500},"required":false,"source_type":"blog","source_url":"https://...","scene":"リビング収納","tags":["定番","公式推奨"]}
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 11d ago First seen · 122 lines · 28 tokens per session scan A 07025053bc32
research-mode is a cursor rule published in the GitHub repository ONE8943/ai-furniture-hub (1 stars, last pushed 4mo ago), licensed MIT. It adds 28 tokens to every session and 1,254 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other cursor rules, from other repositories
about-codebase
General information about the Lineai MCP Server codebase and its purpose.
toc-readme
When updating README.md, the table of contents should contain only the second-level headings below it.
runwhen-infrastructure
Infrastructure conventions for scripts targeting Kubernetes and databases via RunWhen runners.
cursor_rules
Guidelines for creating and maintaining Cursor rules to ensure consistency and effectiveness.
validate-design-early
Ask for design approval before implementing 100+ lines.
ponytail
Ponytail, lazy senior dev mode. Always pick the simplest solution that works.