← 返回 Skills 市场
chenchen913

local-skill-manager

作者 ChenChen · GitHub ↗ · v1.0.1 · MIT-0
cross-platform ✓ 安全检测通过
602
总下载
0
收藏
1
当前安装
2
版本数
在 OpenClaw 中安装
/install local-skill-manager
功能描述
本地 Skill 目录的集中管理工具("Skill 管家")。当用户想要: (1) 列出或查询已安装的 Skill(系统级或用户级); (2) 检查各 Skill 的版本信息; (3) 使用标准模板快速创建新的本地 Skill; (4) 删除或移除不再需要的 Skill。 请使用此技能。
使用说明 (SKILL.md)

本地 Skill 管家

本技能是 skills/ 目录的集中管理工具,提供查询、监控、创建和删除 Skill 的能力,帮助你保持开发环境的整洁有序。

前置依赖

使用本技能前,请确认已安装依赖:

pip install -r requirements.txt

requirements.txt 包含:pyyaml(用于解析 SKILL.md 的 YAML frontmatter)


功能说明

1. 查询与列出 Skill

列出所有已安装的 Skill(含系统级和用户级),显示名称、版本和描述。

触发示例:

  • "列出我的 Skill"
  • "我有哪些 Skill"
  • "查询所有 Skill"

2. 版本检查

检查每个 Skill 的版本号,确认是否为最新版本。

触发示例:

  • "检查 Skill 版本"
  • "我的 Skill 是什么版本"

3. 创建新 Skill

使用标准目录结构(scripts/references/assets/)和 SKILL.md 模板快速脚手架一个新 Skill。

触发示例:

  • "创建一个名为 'data-analyzer' 的新 Skill"
  • "新建一个 Skill"

4. 删除 Skill

安全地移除一个 Skill 目录,内置多层安全保护:路径名格式校验、受保护 Skill 拦截、路径遍历防护、软链接警告、二次确认提示。

触发示例:

  • "删除 'old-test-skill' 这个 Skill"

可用参数:

  • --dry-run:预览将要删除的路径,不实际执行(推荐先用此参数确认)
  • --force:跳过确认提示,直接删除(谨慎使用)

脚本说明

所有脚本位于 scripts/ 目录:

脚本 功能
list_skills.py 扫描并列出所有 Skill
check_versions.py 检查各 Skill 的版本元数据
create_skill.py 初始化新 Skill 的标准目录结构
delete_skill.py 安全移除指定 Skill 目录

执行规则

重要:必须使用以下脚本执行操作,禁止直接使用 lsglob 命令操作 skills/ 目录。

脚本路径结构(从 Skill 根目录相对引用):

# 列出所有 Skill
python scripts/list_skills.py

# 检查版本
python scripts/check_versions.py

# 创建新 Skill(将 skill-name 替换为你的 Skill 名称)
python scripts/create_skill.py "skill-name"

# 删除 Skill(将 skill-name 替换为要删除的 Skill 名称)
# 推荐先用 --dry-run 预览,确认无误后再实际删除
python scripts/delete_skill.py "skill-name" --dry-run
python scripts/delete_skill.py "skill-name"

注意:如果你的平台将 Skills 存放在特定子目录下(如 .trae/skills/),请在命令前加上对应的完整路径前缀,例如:

python .trae/skills/local-skill-manager/scripts/list_skills.py
安全使用建议
This skill appears coherent and focused on managing local Skill directories. Before installing/using it: (1) Run the scripts from the expected location (the scripts compute the skills root as two levels above the scripts/ folder); placing the repo elsewhere will make them operate relative to that location. (2) Use --dry-run and the default confirmation prompt before deleting anything; the delete script includes safety checks but performing a dry-run and backups is prudent. (3) Install requirements (pyyaml) into the correct Python environment (avoid installing into system Python if you don't intend to). (4) Review and run the scripts with a non-privileged user (do not run as root) to reduce blast radius. (5) If you need stricter guarantees, inspect the code yourself or run it in a controlled environment first. Overall there are no network calls or credential accesses in the code.
功能分析
Type: OpenClaw Skill Name: local-skill-manager Version: 1.0.1 The 'local-skill-manager' bundle is a utility for managing local skill directories. It includes scripts for listing, version checking, creating, and deleting skills (scripts/list_skills.py, scripts/create_skill.py, scripts/delete_skill.py). The deletion logic in delete_skill.py includes robust security measures, such as regex-based name validation, a protected skills list, and explicit path-traversal guards using canonicalized paths to ensure operations remain within the designated skills directory. No indicators of data exfiltration, malicious execution, or prompt injection were found.
能力评估
Purpose & Capability
Name/description (manage, list, create, delete local skills) match the included scripts and SKILL.md. All required actions (listing SKILL.md, scaffolding a skill folder, deleting a folder) are implemented and there are no unrelated environment variables, binaries, or external services requested.
Instruction Scope
SKILL.md explicitly instructs using the provided Python scripts for operations and to install pyyaml from requirements.txt. The scripts only read/write files under the computed skills root and parse SKILL.md frontmatter. There are no instructions to read unrelated system files or exfiltrate data.
Install Mechanism
No install mechanism is declared beyond an in-repo requirements.txt (pyyaml). No remote downloads, package installs from arbitrary URLs, or archive extraction are present.
Credentials
The skill requests no environment variables, credentials, or config paths. The code operates on local filesystem paths relative to the script location; this is proportionate to a local skill-manager.
Persistence & Privilege
always is false and the skill does not attempt to modify other skills' configurations or system-wide agent settings. It runs as a normal local utility invoked by the user.
如何使用
  1. 确保已安装 OpenClaw(本地或 Docker 部署)
  2. 在对话框中输入安装命令:/install local-skill-manager
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /local-skill-manager 触发
  4. 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
版本历史
v1.0.1
local-skill-manager v1.0.1 focuses on enhanced Skill 安全移除 support: - 删除 Skill 操作新增多层安全保护,包括路径名校验、受保护 Skill 拦截、路径遍历防护和软链接警告。 - 删除脚本支持 `--dry-run` 预览和 `--force` 强制删除参数。 - 推荐先用 `--dry-run` 参数确认将要删除的内容。 - 其余说明及功能未变。
v1.0.0
- Initial release of local-skill-manager (v1.0.0). - Centralized tool for managing local Skill directories: list, query, check versions, create, and delete Skills. - Provides dedicated scripts for each operation, ensuring safe and standard management (no direct `ls` or `glob` usage). - Supports both system-level and user-level Skills. - Standardizes creation of new Skills with recommended directory structure and metadata template. - Includes command guide and clear script usage instructions.
元数据
Slug local-skill-manager
版本 1.0.1
许可证 MIT-0
累计安装 1
当前安装数 1
历史版本数 2
常见问题

local-skill-manager 是什么?

本地 Skill 目录的集中管理工具("Skill 管家")。当用户想要: (1) 列出或查询已安装的 Skill(系统级或用户级); (2) 检查各 Skill 的版本信息; (3) 使用标准模板快速创建新的本地 Skill; (4) 删除或移除不再需要的 Skill。 请使用此技能。 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 602 次。

如何安装 local-skill-manager?

在 OpenClaw 或 Claude Code 对话框中运行命令「/install local-skill-manager」即可一键安装,无需额外配置。

local-skill-manager 是免费的吗?

是的,local-skill-manager 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。

local-skill-manager 支持哪些平台?

local-skill-manager 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。

谁开发了 local-skill-manager?

由 ChenChen(@chenchen913)开发并维护,当前版本 v1.0.1。

💬 留言讨论