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/numekudi/engineering-ethics/ethics-auditnpx skills add numekudi/engineering-ethics --skill ethics-auditgit clone --depth 1 https://github.com/numekudi/engineering-ethicsWrote 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/numekudi/engineering-ethics/ethics-audit)<a href="https://agentmods.dev/skills/numekudi/engineering-ethics/ethics-audit"><img src="https://agentmods.dev/badge/skills/numekudi/engineering-ethics/ethics-audit.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 | $0.00147 | $0.01897 |
| Opus 5 | $0.00073 | $0.00949 |
| Sonnet 5 | $0.00029 | $0.00379 |
| Haiku 4.5 | $0.00015 | $0.00190 |
Grade A, and why
ethics-audit 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.
How it starts
The opening of the file, as written. The whole thing — 99 lines — stays where its author put it; the contents beside it link to each section on GitHub.
技術者倫理 監査ハーネス(ethics-audit)
リポジトリが技術者倫理(IEEE-CS/ACM ソフトウェア工学倫理綱領 第5.2版)に本当に則して いるかを、実コードを読んだ上で・具体項目に照らして・敵対的に検証して採点する。
思想(なぜこの手順か)
RINRI ライセンスの Golden Subroutine が戒めるとおり、倫理を実質の裏付けなく掲げる行為 (ethical grandstanding)は監査の敵である。本ハーネスは次を鉄則とする:
- コードを読んでから判断する — 印象ではなく
file:lineの証跡に基づく。 - 合否を出せる — バッジのラバースタンプではなく、合格も不適合も出す。
- 誤検知を検証で潰す — 各所見を反証エージェントが攻撃し、生き残りだけを確定する。
- 公共の利益が中心 — 迷ったら「最も権限の弱いユーザーがどう害されるか」で判断する。
- 自己評価の道具 — 他者糾弾ではなく、自リポの改善のために使う。
CLI(決定論的スクリプト)
このスキルに同梱の lib/cli.js(Node.js・実行時依存ゼロ)を使う。以降 <SKILL_DIR> は
この SKILL.md があるディレクトリ。実行例:
node <SKILL_DIR>/lib/cli.js scan <repo-path> # ダークパターン等の候補検出
node <SKILL_DIR>/lib/cli.js principles --principle N # 原則Nの項目を供給
node <SKILL_DIR>/lib/cli.js principles --item N.MM # 単一項目を引く
node <SKILL_DIR>/lib/cli.js score --input findings.json # 確定所見を採点
node <SKILL_DIR>/lib/cli.js tweet --input audit.json --label <名前> # 共有下書き生成
出力は既定で JSON(--format text で人間可読)。エラーは stderr に JSON、exit 1。
監査フロー
番号順に、決定論部分は CLI、判断部分はエージェントに委ねる。対象パスが未指定なら カレントリポジトリを対象にする。
Step 0 — スコープ確定と候補収集(まずコードを読む)
- 対象リポのファイル構成を
Glob/Readで把握する。 node <SKILL_DIR>/lib/cli.js scan <repo>を実行し、候補(ScanCandidate[])を得る。 これは判定ではなく候補。以降の判断の出発点にすぎない。scanはルール定義ファイル自体にも一致しうる(自己言及の誤検知)。対象リポの実体で 判断すること。
Step 1 — ファンアウト評価(8原則を分担)
Agent ツールで評価エージェントを 4〜5 体、並行起動する。各エージェントに:
- 担当原則の項目を
principles --principle N --format textで渡す(推奨割当: A=原則1,3 / B=原則2,6 / C=原則4,5 / D=原則7,8)。 - 別途 ダークパターン・ハンターを1体、
02-dark-patterns.mdを渡して起動する。 - Step 0 の候補と、実コードを読む指示を与える。
- 評価観点は
01-principles.md、深刻度基準は03-scoring-rubric.mdを参照させる。
各エージェントは所見を findings JSON で返す。1件の形:
{"itemId":"3.12","severity":"high","file":"src/log.ts","line":5,
"description":"…","evidence":"該当コードの引用","recommendation":"…","confirmed":false}
この段階では confirmed:false(未検証)。証跡(file:line と evidence)を必須とし、
抽象的な非難は捨てさせる。
Step 2 — 敵対的検証(反-grandstanding の中核)
04-adversarial-verification.md に従い、各所見に反証エージェントを割り当てる。
「実コード・文脈を読んだか? これは grandstanding ではないか? 不確実なら反証」を
基準に REFUTE を試みる。生き残った所見のみ confirmed:true にする。棄却理由も記録する。
What ships with it
15 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- 01-principles.md 5.0 KB
- 02-dark-patterns.md 3.7 KB
- 03-scoring-rubric.md 2.6 KB
- 04-adversarial-verification.md 3.1 KB
- 05-output-format.md 3.1 KB
- data/se-code.json 33 KB
- lib/cli.js 6.1 KB runs code
- lib/commands/principles.js 2.0 KB runs code
- lib/commands/scan.js 7.1 KB runs code
- lib/commands/score.js 4.1 KB runs code
- lib/commands/tweet.js 2.2 KB runs code
- lib/ethics.js 999 B runs code
- lib/schema.js 496 B runs code
- lib/validate.js 7.7 KB runs code
- package.json 250 B
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 · 99 lines · 147 tokens per session scan A 776e6043eaae
ethics-audit is a skill published in the GitHub repository numekudi/engineering-ethics (6 stars, last pushed 1mo ago), licensed MIT. It adds 147 tokens to every session and 1,897 once invoked, about $0.0007 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
specification-writing
Write the full patent specification from claims and invention disclosure. Use when user says "撰写说明书", "write specification", "写说明书", "patent description", or wants to draft the complete patent specification.
regulatory-research-fallback
Fallback workflow for regulatory research when web extraction tools fail on government PDFs.
x-scorecard
OpenSSF Scorecard for assessing open source project security. Check security best practices and compliance. Dependency: This is an x-cmd module. Install x-cmd first (see x-cmd skill for installation options). see x-cmd skill for installation.
memstack-business-gdpr
Use this skill when the user says 'GDPR', 'data protection', 'privacy compliance', 'DPA', 'DSAR', 'data subject request', 'cookie consent', 'privacy audit', 'CCPA', or asks 'do I need GDPR for this repo'. Scans the repository to detect what personal data is collected, classifies sensitivity, determines whether GDPR…
external-purchases
US web checkout via the StoreKit External Purchase Link entitlement — currently 0% Apple commission (litigation ongoing), how to ship it safely, and how to architect for a commission flip so a future ruling is a config change, not a rewrite. Use when adding external purchase links, weighing web checkout vs IAP, or…
nda-review
Use when the user uploads or pastes a non-disclosure agreement and asks for review, redline, risk assessment, or a recommendation on whether to sign. Identifies missing standard protections, one-sided or unusual provisions, and operational issues; produces a structured report with severity ratings and citations to…