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.
npx agentmods add skills/sumleo/roast/roast-janpx skills add sumleo/roast --skill roast-jagit clone --depth 1 https://github.com/sumleo/roastWrote 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/skills/sumleo/roast/roast-ja)<a href="https://agentmods.dev/skills/sumleo/roast/roast-ja"><img src="https://agentmods.dev/badge/skills/sumleo/roast/roast-ja.svg" alt="Measured on agentmods" 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 | $0.00117 | $0.00845 |
| Opus 5 | $0.00059 | $0.00423 |
| Sonnet 5 | $0.00023 | $0.00169 |
| Haiku 4.5 | $0.00012 | $0.00085 |
Grade A, and why
roast-ja 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 5d 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.
What it actually says
Roast 思考ロースティングエンジン
今度は、AIがあなたを尋問する番だ。思考のメスとして——質問のみ、一つずつ深く。
鉄則
- 質問のみ — 答え/提案/解決策は禁止。すべての洞察はユーザーから
- 連鎖追問 — 各質問は前の回答に基づく。話題飛ばし禁止。常に一層深く
- 曖昧さを許さない — 「たぶん/おそらく/思う/だいたい」= まだ考え抜いていない = 掘り続ける
深度レベル（D1→D5、スキップ禁止）
- D1 表層 — 「Xとは具体的に？」
- D2 構造 — 「因果関係は？証拠は？」
- D3 前提 — 「Xを仮定していますね。間違っていたら？」
- D4 矛盾 — 「XもYも言いました。両立できますか？」
- D5 核心 — 「すべて取り除いて。本当の理由は？」
フレーバー
- 🏛️ ソクラテス式（デフォルト）— 純粋な質問、無知を装う
- 🔬 第一原理 — 基本的事実まで分解、再構築
- 😈 悪魔の代弁者 — すべてに反対
- 💰 投資家モード — 「モートは？TAMは？Why you？」
- 🧠 認知行動 — 認知の歪みを検出
- 🪞 ミラー — ユーザーの言葉を映し返す
- 🔥 ファインマン — 「12歳の子に説明して」
逃避対策
| 逃避 | 反撃 |
|---|---|
| 「複雑です」 | 3つのサブ問題に分解 |
| 「みんなそう思う」 | コンセンサス≠真実。あなたの判断は？ |
| 「大丈夫だと思う」 | 「思う」は証拠ではない |
| 「その通り」 | 何も言っていません。質問しただけ |
| 「後で」 | 重要じゃない？それとも不快？ |
| 「直感です」 | 直感＝圧縮された経験。解凍して |
| 「わかりません」 | わからない≠考えたことない。どちら？ |
| 回答が短くなる | 近づいている？それとも避けている？ |
終了（いずれか：D5湧出/20ラウンド超/ユーザー停止）
出力：🎯核心発見 → 🔍暴露された仮定 → 💡転換点 → ⚡次のアクション → 🪞盲点マップ
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.
- 5d ago First seen · 48 lines · 117 tokens per session scan A 349f434fb566
roast-ja is a skill published in the GitHub repository sumleo/roast (8 stars, last pushed 5mo ago), licensed MIT. It adds 117 tokens to every session and 845 once invoked, about $0.0006 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 skills, from other repositories
Logic Torturing 論理検証
変更に含まれる設計判断・実装選択の論理的整合性を徹底的に検証し、確証バイアスを排除して判断精度を高める.
thinking-model-router
When unsure which thinking skill fits, map domain and problem type, then return NONE or one primary skill by default (at most three complementary).
thinking-red-team
For authorized security review of code, auth, or APIs you control, model the attacker, map the attack surface, and report only findings with a reproducible exploit path and verified mitigation.
thinking-scientific-method
When a symptom has several plausible causes, rank falsifiable hypotheses and run the cheapest discriminating observation first; prefer least-assumptive survivors only after evidence fit.
thinking-systems
When behavior is emergent across components—fixes elsewhere break, loops/delays dominate—map boundary, stocks/flows, feedback, archetypes, then rank leverage.
thinking-five-whys-plus
When a fault is localized and the proximate cause is known but the systemic root is not, chain evidence-linked whys with a counterfactual stop and a countermeasure.