home / skills / masayan1126 / masayan-uni-code-plugins / skill-scanner

skill-scanner skill

safe

/plugins/skill-scanner/skills/skill-scanner

This skill scans registered Claude agent skills on Mac and lists them in read-only mode for safe review.

npx playbooks add skill masayan1126/masayan-uni-code-plugins --skill skill-scanner

Review the files below or copy the command above to add this skill to your agents.

Files (2)

SKILL.md

1.0 KB

---
name: skill-scanner
description: Macで登録済みClaude agent skillsをスキャンし一覧表示。「スキルを調べて」「登録済みスキル一覧」などで使用。読み取り専用で安全に実行。
---

# Skill Scanner

登録されているClaude agent skillsをスキャンし一覧表示する（読み取り専用）。

## 使用方法

詳細は `scripts/scan_skills.py --help` を参照。

```bash
# 基本スキャン（ユーザーレベルのみ）
python3 scripts/scan_skills.py

# 指定ディレクトリ以下のプロジェクトをスキャン（要ユーザー確認）
python3 scripts/scan_skills.py --projects-dir ~/git
python3 scripts/scan_skills.py --projects-dir ~/work

# 複数ディレクトリを同時にスキャン
python3 scripts/scan_skills.py --projects-dir ~/git --projects-dir ~/work

# JSON形式で出力
python3 scripts/scan_skills.py --json
```

**重要**: プロジェクトディレクトリをスキャンする際は必ずユーザーに確認してから実行。

Overview

This skill scans and lists Claude agent skills registered on a Mac, providing a safe read-only inventory. It helps you discover installed skills, export results as JSON, and optionally scan specified project directories with user confirmation.

How this skill works

The scanner inspects system and user-level locations where Claude agent skills are registered and aggregates metadata into a readable list. It can optionally traverse one or more project directories to detect skills in development, but always prompts for user confirmation before scanning external project folders. Output can be printed or emitted as JSON for integration with other tools.

When to use it

Check which Claude agent skills are currently installed on your Mac.
Audit skills before upgrades or system changes to ensure compatibility.
Build an inventory for documentation or development handoffs.
Quickly detect skills present in one or more project directories.
Export skill lists to JSON for tooling or automation.

Best practices

Run the basic scan first (user-level only) to get a quick, safe inventory.
Always confirm prompts before scanning project directories to avoid unexpected traversal.
Use JSON output when integrating results with other scripts or dashboards.
Limit scanning to trusted project paths to reduce noise and speed up results.
Review discovered metadata to verify versions and source locations.

Example use cases

Perform a user-level scan to list installed skills before installing new ones.
Scan ~/git and ~/work directories to find skills under development (confirm before running).
Generate JSON output to feed into a CI job that verifies required skills are present.
Create a consolidated inventory for a teammate when onboarding to a project.

FAQ

Is the scan safe to run?

Yes. The scanner runs in read-only mode and does not modify files. It prompts for confirmation before scanning any project directories.

How do I get machine-readable results?

Use the --json option to emit results in JSON format suitable for further processing.