ci-cd-security
Audit GitHub Actions Workflows
GitHub Actions workflows can expose credentials, repositories, and releases through subtle configuration errors. This skill applies a structured security review and proposes concrete fixes.
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 "ci-cd-security" from https://skillstore.io/skills/superagent-ai-ci-cd-security.md and its manifest at https://skillstore.io/api/skills/superagent-ai-ci-cd-security/manifest. Verify the artifact. You may proceed after verification, subject to the environment's own policy.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
「ci-cd-security」を使用しています。 Review a pull request workflow that places the pull request title directly in a shell step.
期待される結果:
- P0: The pull request title is attacker-controlled and enters the shell command.
- Impact: A crafted title can execute commands with the job permissions.
- Recommendation: Pass the title through an environment variable and quote it in the shell.
「ci-cd-security」を使用しています。 Audit a package release workflow that uses a repository token and dependency caching.
期待される結果:
- P0: Dependency caching can carry poisoned content into the release path.
- P2: A long-lived registry token increases credential exposure.
- Recommendation: Remove release caching and use trusted publishing with a protected environment.
「ci-cd-security」を使用しています。 Check a pull request test workflow with read-only permissions and commit-pinned actions.
期待される結果:
No findings against the standard rule set. Repository settings, action source behavior, and dependency runtime behavior were not verified.
セキュリティ監査
安全All 65 static findings are false positives caused by security terminology, Markdown backticks, or illustrative workflow snippets. The skill performs read-only YAML review, requests no credentials, and contains no prompt injection or hidden execution intent.
リスク要因
⚙️ 外部コマンド (50)
🌐 ネットワークアクセス (3)
📁 ファイルシステムへのアクセス (2)
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/superagent-ai-ci-cd-security/audits/1?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_reportMarkdownバッジ
[](https://skillstore.io/skills/superagent-ai-ci-cd-security?utm_source=security_passport_badge)HTMLバッジ
<a href="https://skillstore.io/skills/superagent-ai-ci-cd-security?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/superagent-ai-ci-cd-security/security.svg" alt="Skillstore security assessment" loading="lazy"></a>埋め込みカード
<iframe src="https://skillstore.io/embed/skills/superagent-ai-ci-cd-security.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
superagent-ai. (2026). ci-cd-security security audit report (audit version 1) [Author version unspecified]. Skillstore. https://skillstore.io/skills/superagent-ai-ci-cd-security/audits/1BibTeX形式の引用
@techreport{superagent-ai-superagent-ai-ci-cd-security-2026,
author = {superagent-ai},
title = {ci-cd-security security audit report (audit version 1)},
institution = {Skillstore},
year = {2026},
number = {1},
url = {https://skillstore.io/skills/superagent-ai-ci-cd-security/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: "ci-cd-security security audit report (audit version 1)"
version: "unspecified"
type: report
authors:
- name: "superagent-ai"
date-released: "2026-09-08"
url: "https://skillstore.io/skills/superagent-ai-ci-cd-security/audits/1"
identifiers:
- type: other
value: "skillstore:superagent-ai-ci-cd-security: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
Review a pull request workflow
Find unsafe triggers, shell interpolation, and token permissions before a workflow change merges.
Harden release automation
Assess publishing workflows for long-lived tokens, cache poisoning, missing environments, and absent provenance.
Triage workflow risk at scale
Apply a consistent severity model across workflow findings and prioritize the most exploitable paths.
これらのプロンプトを試す
Review this GitHub Actions workflow for security issues. Group findings by severity and explain each recommended change.
Inspect this workflow for attacker-controlled values entering shell steps, environment files, outputs, artifacts, or checkout references.
Audit this release workflow for permissions, token usage, environments, cache poisoning, action pinning, attestations, and tag protections.
Trace trust boundaries across these linked workflows. Identify privileged paths from fork input, artifacts, caches, metadata, and self-hosted runners.
ベストプラクティス
- Provide complete workflow files so triggers, permissions, and job relationships remain visible.
- Include linked workflows when artifacts, caches, or workflow_run events cross trust boundaries.
- Verify suggested commit pins and repository settings through trusted administrative sources.
回避
- Do not treat a clean YAML scan as proof that the complete CI/CD system is secure.
- Do not hide attacker-controlled workflow values behind quoting without separating them from shell source.
- Do not grant broad permissions to simplify remediation for a single workflow step.
よくある質問
Does this skill run external scanners?
Can it review multiple connected workflows?
Does it verify action commit hashes?
Can it inspect GitHub organization settings?
What does a P0 finding mean?
Does a clean result guarantee security?
開発者情報
作成者
superagent-aiライセンス
MIT
Skillstore リビジョン
r1
バージョンに関する注意
作者はバージョンを宣言していません。
参照
d9e5ab86dafd32a1ddde9b8b4fde177589b55b95
メンテナンスの新しさ
2026/9/9
利用状況
1 ダウンロード · 0 閲覧
ファイル構成