whatileaked
Find Credentials Leaked in AI Coding Histories
AI coding histories can retain credentials after they were shared with a model provider. This skill finds affected records while keeping reported secrets masked.
インストール前に停止して確認を求めてください。
計画を確認し、ファイルを変更する前にユーザーの明示的な同意を得てください。
自分のエージェントでインストール
このリクエストをエージェントにコピーしてください。正規の Skill ページとマニフェストが含まれています。
Review the Skillstore skill "whatileaked" from https://skillstore.io/skills/selan-ai-whatileaked.md and its manifest at https://skillstore.io/api/skills/selan-ai-whatileaked/manifest. Verify the artifact. Stop and obtain explicit user consent before installing or changing files.エージェントは引き続き計画を提示し、セキュリティポリシーで必要な確認を求める必要があります。
エージェントが読めるリソース
AI エージェント、クローラー、スクリプトがページ全体ではなく整理されたコンテキストを必要とする場合は、これらのリンクを使ってください。
テストする
「whatileaked」を使用しています。 Scan my coding-agent histories for leaked credentials.
期待される結果:
The scan found two fingerprints across three files. Rotate the likely production credential before considering local cleanup.
「whatileaked」を使用しています。 Are these two findings the same credential?
期待される結果:
Yes. The matching fingerprint indicates one credential appeared in both locations, although the secret itself remains masked.
「whatileaked」を使用しています。 Help me interpret a masked test-key finding.
期待される結果:
- The context suggests a test fixture rather than an active credential.
- Confirm the source file and deployment use before dismissing it.
- Rotate the credential if its status remains uncertain.
セキュリティ監査
高リスクAll four static findings are false positives caused by Markdown prose and masked examples, not executable Ruby or environment access. The skill still introduces real risks by installing an external package and offering an irreversible wipe operation.
確認済みのセキュリティ上の懸念 (2)
リスク要因
⚙️ 外部コマンド (2)
🔑 環境変数 (1)
このレポートを共有・引用
バージョン付き評価レポート、中立的なバッジ、埋め込みカード、引用を共有できます。Skillstore は証拠を報告しますが、この Skill が安全かどうかは判断しません。
レポートリンクをコピー
https://skillstore.io/skills/selan-ai-whatileaked/audits/1?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_reportMarkdownバッジ
[](https://skillstore.io/skills/selan-ai-whatileaked?utm_source=security_passport_badge)HTMLバッジ
<a href="https://skillstore.io/skills/selan-ai-whatileaked?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/selan-ai-whatileaked/security.svg" alt="Skillstore security assessment" loading="lazy"></a>埋め込みカード
<iframe src="https://skillstore.io/embed/skills/selan-ai-whatileaked.html" title="Skillstore Security Assessment" sandbox="allow-popups allow-popups-to-escape-sandbox" loading="lazy" referrerpolicy="no-referrer" width="420" height="180"></iframe>学術引用 (APA · BibTeX · CFF)
APA形式の引用
selan-ai. (2026). whatileaked security audit report (audit version 1) [Author version unspecified]. Skillstore. https://skillstore.io/skills/selan-ai-whatileaked/audits/1BibTeX形式の引用
@techreport{selan-ai-selan-ai-whatileaked-2026,
author = {selan-ai},
title = {whatileaked security audit report (audit version 1)},
institution = {Skillstore},
year = {2026},
number = {1},
url = {https://skillstore.io/skills/selan-ai-whatileaked/audits/1},
note = {Author version unspecified}
}CITATION.cff
cff-version: 1.2.0
message: "If you use this Skill, cite its author and this versioned security audit report."
title: "whatileaked security audit report (audit version 1)"
version: "unspecified"
type: report
authors:
- name: "selan-ai"
date-released: "2026-08-30"
url: "https://skillstore.io/skills/selan-ai-whatileaked/audits/1"
identifiers:
- type: other
value: "skillstore:selan-ai-whatileaked:audit:1"
description: "Skillstore immutable audit report identifier"
Skillstore スコア
このスコアの理由 証拠の信頼度: 中作成できるもの
Audit a developer workstation
Find credentials retained in local coding-agent histories before sharing or transferring a machine.
Investigate a credential exposure
Identify affected projects, files, and repeated appearances during incident response.
Prepare managed devices for reassignment
Review agent history findings and plan credential rotation before approved cleanup.
これらのプロンプトを試す
Check whether whatileaked is available, then scan my local supported agent histories. Do not install anything without asking.
Review the whatileaked results and explain which findings appear real, which may be fixtures, and which credentials need rotation.
Use the scan results to create a prioritized rotation and cleanup plan. Group repeated fingerprints as the same credential.
Review every file that the wipe operation would change. Proceed only after my explicit request and never answer its confirmation prompt automatically.
ベストプラクティス
- Rotate confirmed credentials before deleting local copies.
- Review masked context and repeated fingerprints before classifying findings.
- Require explicit approval for installation and every destructive cleanup action.
回避
- Do not treat local deletion as credential revocation.
- Do not paste raw credentials into follow-up prompts or reports.
- Do not automate or bypass the wipe confirmation prompt.