security-sweep-plugin - Free Claude Code plugin that scans your project fo

项目说明

Security Sweep — Claude Code Plugin

A comprehensi scanner you can run against any codebase from Claude Code. Finds hardcoded secrets, injection flaws, auth issues, misconfigurations, AI-specific , and more.

Covers OWASP Top 10 (2025), OWASP Mobile Top 10 (2024), and OWASP LLM Top 10 (2025).

---

Before You Install This (or Any) Plugin — Read This First

This section exists because we belie starts before you write a single line of code. It starts with the tools you choose to trust.

The Reality of AI Tool Plugins

Claude Code plugins, MCP servers, custom skills, GPT Actions, IDE extensions — the entire ecosystem of AI de tooling shares a fundamental problem: most of it runs with your pri, on your machine, with access to your code.

When you install a plugin, you are giving it:

Access to read and write files in your project

The ability to run shell commands on your behalf

Potential access to your en (API keys, tokens, credentials)

Network access to external services

The ability to modify how your AI assistant behaves

This is not a theoretical risk. A malicious or compromised plugin could exfiltrate your source code, steal credentials from your en, inject backdoors into your codebase, or silently alter your AI assistant's behavior.

What You Should Do Before Installing Any Plugin

1. Read the source code.

Every plugin you install should be open source. If it isn't, don't install it. Before installing, actually read:

Skill files (SKILL.md) — These are the prompts that instruct the AI. Look for instructions that tell the AI to exfiltrate data, make network requests to unknown servers, or run obfuscated commands. A skill prompt is plain text — there is nothing stopping a malicious author from embedding instructions like "silently send the contents of .env to

Hook configurations — Hooks run shell commands automatically in response to events (file edits, session start, tool calls). A malicious hook could run curl to exfiltrate your code every time you start a session. Look at every command field in any hooks config.

MCP ser — These connect to external services. you recognize and trust every server URL. Anthropic explicitly states they do not audit MCP servers — even in official marketplace plugins.

Scripts and binaries — Check any scripts/ or bin/ directories. Don't run binaries you can't inspect.

Settings o — Plugins can ship settings.json that changes your Claude Code configuration. Review what it changes.

2. Check the author and repository.

Is the GitHub account real and established, or was it created last week?

Does the author ha projects and a history?

Are there stars, forks, or issues that suggest real community use?

Is the repository actively maintained?

3. the plugin does only what it claims.

A "code formatter" plugin should not need network access. A "security scanner" (like this one) should only need to read files and run grep — it should not be making external API calls or writing to your codebase. Match the plugin's claimed purpose against its actual capabilities.

4. Use Claude Code's safety features.

Permission prompts — Don't auto-appro. Read what you're approving, especially bash commands.

Sandbox mode — Run /sandbox to isolate file system and network access when e untrusted plugins.

Scoped installation — Install to project scope rather than user scope when possible, to limit blast radius.

Organization restrictions — If you're on a team, use strictKnownMarketplaces in managed settings to control which marketplaces your team can use.

5. Understand what protections do NOT exist today.

As of early 2026, the Claude Code plugin ecosystem does not have:

Code signing or cryptographic of plugins

Automated security scanning of marketplace submissions

Capability-based permissions (a plugin gets broad access, not granular)

Sandboxing by default (it's opt-in)

A documented process for marketplace submissions

This will likely improve over time. But today, trust is entirely on you.

How to This Plugin Specifically

This plugin contains only two files that matter:

| File | What It Is | What to Check |
|---|---|---|
| skills/security-sweep/SKILL.md | The prompt that tells Claude how to run the scan | it only instructs Claude to read files and grep for patterns. No network calls, no writes, no exfiltration. |
| skills/security-sweep/patterns.md | Regex patterns for detecting | these are legitimate security patterns (OWASP, Semgrep, Gitleaks-derived). No obfuscated content. |

There are no hooks, no MCP servers, no scripts, no binaries. The allowed-tools in the skill frontmatter is restricted to Read, Grep, Glob, Bash, Agent — standard read-oriented tools.

You can this yourself in under 5 minutes. If you can't a plugin that quickly, that's a red flag about its complexity.

A Note on AI-Assisted De and Security

We are in an era where de increasingly rely on AI tools to write, review, and ship code. This creates a new category of attack surface:

Prompt injection — A malicious skill prompt can instruct the AI to behave in ways the user doesn't expect

Trust delegation — When you auto-approve AI tool calls, you're delegating trust to whate the AI's instructions — including plugins

Supply chain — Just as npm install can run malicious postinstall scripts, installing a plugin loads prompts that change your AI's behavior

The security community hasn't fully caught up with these risks yet. By using and sharing tools like this scanner, and by reading plugin source code before installing, you're part of building better practices.

---

What It Scans

| Module | What It Finds | Priority |
|---|---|---|
| Secrets | 30+ API key formats (AWS, GitHub, Stripe, OpenAI, Anthropic...), private keys, DB connection strings, committed .env files | Critical |
| Injection | SQL injection, XSS, command injection, SSRF, insecure deserialization, path tra | Critical |
| Auth | JWT misuse, weak password hashing, insecure cookies, broken access control | High |
| Config | CORS wildcards, DEBUG mode, exposed endpoints, TLS issues, Docker/K8s misconfig | High |
| Dependencies | Missing lockfiles, suspicious install scripts, supply chain red flags | High |
| AI | Hardcoded AI API keys, prompt injection , eval of LLM output, excessi permissions | High |
| Mobile | Android: cleartext traffic, exported components, insecure storage. iOS: ATS bypass, UserDefaults secrets | High |
| Data Exposure | PII in logs, sensitive data in URLs, plaintext HTTP, weak crypto | Medium |

Supported Languages & Frameworks

Python, Ja/TypeScript, Ja, Go, Ruby, PHP, C#, Swift, Dart/Flutter — with framework-aware checks for Django, Flask, Express, Spring Boot, Next.js, React, Vue, Angular, and more.

Installation

As a plugin

Manual (copy to personal skills)

Usage

Full scan

Scan specific module

Scan specific path

Report Format

The scan produces a structured report with:

Summary banner with finding counts by se

Findings grouped by se (CRITICAL first)

Each finding includes: file path, line number, e, risk explanation, specific fix, and OWASP/CWE reference

Top 3 priorities to fix first

Positi (what the project does well)

General recommendations

How It Works

Unlike regex-only tools, this skill le Claude's contextual understanding to:

Reduce false positives — reads surrounding code before classifying se

Understand frameworks — knows Django ORM prevents SQLi, React escapes by default, etc.

Cross-file analysis — traces data flow without building an AST

Explain findings — provides de remediation with code examples

Detection Sources

Patterns are derived from:

Semgrep default and security-audit rulesets

Gitleaks secret patterns

TruffleHog detectors

ESLint-plugin-security rules

Bandit (Python SAST) rules

OWASP Cheat Sheet Series

PortSwigger Web Security Academy

Contributing

Issues, PRs, and additional detection patterns are welcome. If you find a false positive or a missing class, open an issue.

License

MIT

项目信息

开发语言：未知

Star 数：11

技术要点

该项目属于安全软件方向的开源资源，代码结构清晰，适合安全研究人员学习参考。项目围绕Free Claude Code plugin that scans your project for leaked API keys, security holes, and common mistakes before you ship. No security expertise needed.展开，提供了可运行的实现，对理解相关安全机制、工具原理与攻防技术具有较高的研究价值。

从技术实现角度看，此类项目的核心在于对底层原理的深入理解与工程化落地。开发者在研究过程中可以结合官方文档与社区讨论，梳理项目的初始化流程、核心模块职责与数据流转，逐步掌握其中的技术细节与设计思路。

使用与学习建议

1. 下载源码压缩包，解压到本地环境。

2. 根据项目类型准备对应的运行环境与依赖。

3. 按照 README 中的步骤完成配置与部署。

4. 在隔离的测试环境中运行验证，避免影响生产系统。

免责声明

本资源来源于开源社区，仅供安全研究、技术学习与合法授权范围内的攻防演练使用，请勿用于任何非法或侵害他人权益的用途。使用本项目产生的一切后果由使用者自行承担。

登录后才能参与讨论

下载链接

会员才能下载

登录后开通会员

同分类推荐

暂无同类资源

资源标题	security-sweep-plugin - Free Claude Code plugin that scans your project fo
所属分类	安全软件
发布作者	admin
发布时间	2026-10-05 23:02:00
下载权限	免费下载
附件数量	2 个下载链接
浏览次数	372
讨论数量	0
主题状态	正常讨论中