← 返回 Skills 市场
twunkdean-cloud

Custom Commands

作者 twunkdean-cloud · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ⚠ suspicious
114
总下载
0
收藏
1
当前安装
1
版本数
在 OpenClaw 中安装
/install custom-commands
功能描述
Create and manage custom commands like backup, sync, clean, generate, and audit to automate file tasks and content workflows efficiently.
使用说明 (SKILL.md)
# custom-commands/SKILL.md

## Custom Commands

Define and manage custom commands like `backup`, `sync`, `clean`, `generate`, and `audit` for streamlined workflows.

### Commands

- `backup [path]` - Backs up files/folders to a specified location (default: cloud storage)
- `sync [source] [destination]` - Synchronizes files between locations
- `clean [pattern]` - Deletes temporary files matching a pattern
- `generate [type]` - Generate content (articles, scripts, outlines) based on prompts
- `audit [scope]` - Run security or workflow audits (e.g., "Audit content approval process")

### Integration

- Voice command support: "Hey Assistant, backup my documents" or "Write a 500-word article on AI ethics"
- Auto-suggest these commands in relevant contexts

### Memory Optimization

- **Rule 1**: Save only high-impact decisions (e.g., "Approved backup schedule: daily at 2 AM", "Adjusted content approval workflow")
- **Rule 2**: Redact IPs, credentials, or internal URLs before storing (e.g., "Client email: [REDACTED]", "Internal URL: [REDACTED]")
- **Rule 3**: Archive completed tasks to `memory/archived/` with dated filenames (e.g., `memory/archived/2026-03-18_content-review.md`)

安全使用建议
This skill's behavior is plausible for a 'custom commands' helper, but it leaves important questions unanswered. Before installing or enabling it: 1) Ask the author where backups go and which cloud provider/API is used; require explicit declarations of any environment variables or credentials that will be needed. 2) Require safe defaults: a dry-run mode, explicit user confirmation for delete/sync/backup actions, and a whitelist of directories the skill may operate on. 3) Insist the skill declare any filesystem paths it will write to (e.g., memory/archived/) and provide a way to change that path. 4) Verify redaction is actually enforced (prefer deterministic redaction before storage or transmission). 5) If you must use it, run first in a restricted/test environment and audit actions/logs closely. If the author cannot clarify these points or provide source code, treat it as higher risk and avoid granting filesystem or cloud credentials.
功能分析
Type: OpenClaw Skill Name: custom-commands Version: 1.0.0 The skill bundle defines a set of administrative and content generation commands (backup, sync, clean, generate, audit) for the OpenClaw agent. The SKILL.md file includes privacy-conscious instructions for memory management, specifically requiring the redaction of credentials, IPs, and internal URLs before storage, and shows no evidence of malicious intent or harmful prompt injection.
能力评估
Purpose & Capability
The description promises file-oriented commands (backup, sync, clean) and content workflows. The SKILL.md explicitly references 'cloud storage' as the default backup target and an on-disk archive path (memory/archived/). Yet the skill declares no required environment variables, no credentials, and no required config paths. Backing up to cloud normally requires credentials/config; archiving implies filesystem writes. This mismatch between stated purpose and declared requirements is concerning.
Instruction Scope
Runtime instructions include destructive and sensitive actions: deleting files by pattern (clean), synchronizing locations (sync), and backing up (defaulting to cloud). They also direct archiving completed tasks into a specific path and give a redaction rule. The instructions are high-level and lack safe-guards (no dry-run, no confirmation step, no whitelist of allowed directories). That gives an agent broad discretion to read, write, delete, or transmit files unless external controls exist.
Install Mechanism
This is an instruction-only skill with no install spec and no code files. That minimizes supply-chain risk (nothing is downloaded or written by install). The static scanner had nothing to analyze because there are no code files.
Credentials
The SKILL.md implies use of cloud storage and filesystem archives but the skill requests no credentials, API keys, or config paths. Reasonable implementations would require at least cloud credentials or an explicit note that the agent will prompt for them. The absence of declared env vars or config paths is disproportionate to the operations described.
Persistence & Privilege
The skill is not always-enabled and is user-invocable (normal). However, it instructs the agent to write archives to memory/archived/, which is a form of persistent storage. That path is not declared in required config paths. Autonomous invocation combined with write/delete operations increases the risk surface if the agent is allowed to act without explicit user confirmation.
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 custom-commands
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /custom-commands 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
v1.0.0
Initial release of custom-commands skill for streamlined workflow automation: - Introduces commands: `backup`, `sync`, `clean`, `generate`, and `audit` - Supports voice commands and auto-suggestions in relevant contexts - Implements memory rules for efficient storage, redaction, and task archiving
元数据
Slug custom-commands
版本 1.0.0
许可证 MIT-0
累计安装 1
当前安装数 1
历史版本数 1
常见问题

Custom Commands 是什么?

Create and manage custom commands like backup, sync, clean, generate, and audit to automate file tasks and content workflows efficiently. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 114 次。

如何安装 Custom Commands?

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

Custom Commands 是免费的吗?

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

Custom Commands 支持哪些平台?

Custom Commands 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 Custom Commands?

由 twunkdean-cloud(@twunkdean-cloud)开发并维护,当前版本 v1.0.0。

💬 留言讨论