dockerfile-scan

スター12

フォーク5

更新日2026年6月23日 07:55

Scan a Dockerfile for insecure build patterns — running as root, unpinned or :latest base images, ADD from remote URLs, piping curl/wget into a shell, hardcoded secrets in ENV/ARG, world-writable chmod 777, and sudo usage. Use when the user asks to "review my Dockerfile", "is this container image secure", "lint my Dockerfile for security", or "why does my image run as root".

インストール

Codex または Claude でインストールこの Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。

ソース

NovaCode37

NovaCode37/claude-security-skills

GitHub リポジトリを開く Creator のリポジトリを見る

ダウンロード

Dockerfile Security Scan

Parses a Dockerfile and flags common security and supply-chain problems, each with a concrete fix. Pure text analysis — no Docker daemon, no network, stdlib only.

When to use this skill

"Review this Dockerfile for security issues."
"Is my container running as root?"
"Lint my Dockerfiles before I push them."

What it checks

Base image pinning — missing tag or :latest (docker-no-tag, docker-latest-tag).
Runs as root — no USER, or final USER is root (docker-root-user).
Remote code execution — curl|wget … | sh (docker-remote-exec).
Remote ADD — ADD http://… without integrity checks (docker-add-remote).
Hardcoded secrets — ENV/ARG with a secret-looking name and a value (docker-hardcoded-secret).
Loose permissions — chmod 777 (docker-chmod-777).
sudo usage in RUN (docker-sudo).

How to run it

# Scan one Dockerfile
python skills/dockerfile-scan/scanner.py path/to/Dockerfile

# Scan every Dockerfile under a directory, as JSON
python skills/dockerfile-scan/scanner.py . --json

Exit codes: 0 clean · 1 findings present · 2 usage/IO error.

Recommended workflow for Claude

Run the scan on the file or repo.
Present findings by severity, each with its one-line fix.
Offer a corrected Dockerfile snippet (pinned base image, non-root USER, COPY instead of remote ADD, secrets moved to runtime).

このリポジトリの他の Skills

同じリポジトリ

cors-auditor

NovaCode37/claude-security-skills

Audit a site's Cross-Origin Resource Sharing (CORS) configuration for misconfigurations — wildcard origin with credentials, reflected arbitrary Origin, the 'null' origin, overly broad allowed methods, and risky credentialed CORS. Use when the user asks to "check my CORS config", "is my API's CORS safe", "test for CORS misconfiguration", or "why can any site call my API".

2026-06-2312

dependency-check

NovaCode37/claude-security-skills

Audit project dependencies for known-vulnerable versions and risky pinning. Parses requirements.txt and package.json, matches a bundled offline advisory DB, optionally queries OSV.dev live, and warns about unpinned versions. Use when the user asks to "check dependencies for vulnerabilities", "audit my requirements.txt / package.json", "scan for vulnerable packages", or "is my dependency tree secure".

2026-06-0812

secret-scanner

NovaCode37/claude-security-skills

Scan a codebase for hardcoded secrets — API keys, tokens, private keys and passwords — using a custom regex + Shannon-entropy engine. Use when the user asks to "find secrets", "check for leaked credentials", "scan for API keys", do a pre-commit secret check, or audit a repo before making it public.

2026-06-0712

prompt-injection-tester

NovaCode37/claude-security-skills

Red-team an LLM application against prompt injection and jailbreaks using a curated, categorized payload library and canary-based detection, then produce a resilience score. Use when the user asks to "test my chatbot for prompt injection", "check if my AI app is jailbreakable", "red-team my LLM", "evaluate prompt-injection defenses", or audit a system prompt's guardrails.

2026-06-0712

jwt-inspector

NovaCode37/claude-security-skills

Decode and security-audit a JSON Web Token — flag alg=none, missing/excessive expiry, symmetric-alg confusion risk, missing claims — and attempt an offline HMAC secret crack against a wordlist to detect weak signing keys. Use when the user asks to "decode this JWT", "is this token secure?", "audit a JWT", or "check if this token uses a weak secret".

2026-06-0712

http-sec-audit

NovaCode37/claude-security-skills

Audit a website's HTTP security headers and cookie flags — CSP, HSTS, X-Frame-Options, X-Content-Type-Options, Referrer-Policy, Permissions-Policy, COOP/COEP, version-leaking banners, and Secure/HttpOnly/SameSite cookies. Use when the user asks to "check my site's security headers", "audit HTTP headers", "is my CSP/HSTS configured right", or "scan a URL for header misconfigs".

2026-06-0712

name	dockerfile-scan
description	Scan a Dockerfile for insecure build patterns — running as root, unpinned or :latest base images, ADD from remote URLs, piping curl/wget into a shell, hardcoded secrets in ENV/ARG, world-writable chmod 777, and sudo usage. Use when the user asks to "review my Dockerfile", "is this container image secure", "lint my Dockerfile for security", or "why does my image run as root".
license	MIT

Dockerfile Security Scan

Parses a Dockerfile and flags common security and supply-chain problems, each with a concrete fix. Pure text analysis — no Docker daemon, no network, stdlib only.

When to use this skill

"Review this Dockerfile for security issues."
"Is my container running as root?"
"Lint my Dockerfiles before I push them."

What it checks

Base image pinning — missing tag or :latest (docker-no-tag, docker-latest-tag).
Runs as root — no USER, or final USER is root (docker-root-user).
Remote code execution — curl|wget … | sh (docker-remote-exec).
Remote ADD — ADD http://… without integrity checks (docker-add-remote).
Hardcoded secrets — ENV/ARG with a secret-looking name and a value (docker-hardcoded-secret).
Loose permissions — chmod 777 (docker-chmod-777).
sudo usage in RUN (docker-sudo).

How to run it

# Scan one Dockerfile
python skills/dockerfile-scan/scanner.py path/to/Dockerfile

# Scan every Dockerfile under a directory, as JSON
python skills/dockerfile-scan/scanner.py . --json

Exit codes: 0 clean · 1 findings present · 2 usage/IO error.

Recommended workflow for Claude

Run the scan on the file or repo.
Present findings by severity, each with its one-line fix.
Offer a corrected Dockerfile snippet (pinned base image, non-root USER, COPY instead of remote ADD, secrets moved to runtime).