Claude Code Plugins

Community-maintained marketplace

Feedback

security-orchestrator

@fujiokayu/dotfiles
0
0

Use this skill when you need to perform comprehensive security vulnerability assessments on a codebase. This skill launches the security-orchestrator agent to conduct systematic security reviews by breaking down the codebase into architectural units and performing deep security analysis.

Install Skill

1Download skill
2Enable skills in Claude

Open claude.ai/settings/capabilities and find the "Skills" section

3Upload to Claude

Click "Upload skill" and select the downloaded ZIP file

Note: Please verify skill by going through its instructions before using it.

SKILL.md

name security-orchestrator
description Use this skill when you need to perform comprehensive security vulnerability assessments on a codebase. This skill launches the security-orchestrator agent to conduct systematic security reviews by breaking down the codebase into architectural units and performing deep security analysis.

Security Orchestration Skill

This skill helps you launch the security-orchestrator agent for comprehensive security vulnerability assessments.

When to Use This Skill

Use this skill when the user requests:

  • Comprehensive security review of their project
  • Security check for API endpoints and controllers
  • Security validation of authentication/authorization code
  • Vulnerability assessment of specific architectural layers
  • Security risk identification before deployment

How to Use

When you detect a security review request, use the Task tool to launch the security-orchestrator agent:

Task tool parameters:
- subagent_type: "security-orchestrator"
- description: Brief description of the security review task (e.g., "Comprehensive security review")
- prompt: Detailed task description including:
  - Scope of the review (full project, specific modules, etc.)
  - Any specific concerns the user mentioned
  - Relevant context about the codebase

Context to Provide to the Agent

When launching the agent, include:

  • Scope: What needs to be reviewed (entire project, specific layers, particular modules)
  • Focus areas: Any specific security concerns mentioned by the user
  • Technology stack: Programming language, frameworks being used (if known)
  • User context: Any recent changes or implementations mentioned

Examples

Example 1: Full Project Security Review

User: "このプロジェクトのセキュリティレビューをお願いします"

Action:

セキュリティレビューを実施します。security-orchestratorエージェントを起動して、
包括的な脆弱性診断を行います。

<Use Task tool with:>
subagent_type: "security-orchestrator"
description: "Comprehensive security review"
prompt: "プロジェクト全体のセキュリティレビューを実施してください。
アーキテクチャ層ごとに分析し、OWASP Top 10に基づく脆弱性を重点的にチェックしてください。"

Example 2: API Security Check

User: "APIエンドポイントとコントローラーのセキュリティチェックをしてください"

Action:

APIとコントローラー層のセキュリティチェックを実施します。

<Use Task tool with:>
subagent_type: "security-orchestrator"
description: "API and controller security check"
prompt: "APIエンドポイントとコントローラー層のセキュリティレビューを実施してください。
インジェクション攻撃、入力検証、認証・認可、エラーハンドリングを重点的にチェックしてください。"

Example 3: Authentication Review

User: "認証周りのコードを実装したので、セキュリティ的に問題ないか確認してほしい"

Action:

認証実装のセキュリティレビューを実施します。

<Use Task tool with:>
subagent_type: "security-orchestrator"
description: "Authentication security review"
prompt: "認証・認可モジュールのセキュリティレビューを実施してください。
パスワード保管、セッション管理、トークン処理、権限昇格リスク、タイミング攻撃への対策を重点的にチェックしてください。"

Communication

  • Acknowledge the user's request in Japanese
  • Briefly explain that you're launching the security-orchestrator agent
  • Launch the agent with appropriate context
  • Wait for the agent to complete and present results to the user in Japanese