← 返回 Skills 市场
solomonneas

Content Scrubber

作者 Solomon Neas · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ✓ 安全检测通过
150
总下载
0
收藏
0
当前安装
1
版本数
在 OpenClaw 中安装
/install content-scrubber
功能描述
OpenClaw plugin that scrubs private infrastructure details from outgoing messages. Regex-based redaction of RFC 1918 IPs, localhost ports, SSH targets, and h...
使用说明 (SKILL.md)

Content Scrubber

An OpenClaw extension plugin that intercepts outgoing messages and redacts private infrastructure details before delivery.

What It Catches

  • RFC 1918 IPv4 addresses: 10.x.x.x, 172.16-31.x.x, 192.168.x.x
  • Loopback addresses: 127.x.x.x
  • localhost with ports: localhost:8080, localhost:3000, etc.
  • SSH/SCP targets: [email protected]:/path
  • Custom hostnames: configurable hostname patterns

How It Works

The plugin registers as a message interceptor. Before any message leaves OpenClaw (Discord, Telegram, Signal, etc.), it runs through regex-based scrubbing rules that replace private details with safe placeholders like [redacted-ip], [redacted-service], [redacted-target].

Rules are deterministic (regex, not LLM), so they're fast, auditable, and never miss edge cases that an LLM scrubber would.

Installation

  1. Copy the plugin files to your OpenClaw extensions directory:

    ~/.openclaw/extensions/content-scrubber/
    ├── index.ts
    ├── openclaw.plugin.json
    └── package.json
    
  2. Add to your openclaw.json plugins config:

    {
      "plugins": {
        "entries": {
          "content-scrubber": {
            "enabled": true,
            "config": {
              "dryRun": false,
              "allowedRecipients": []
            }
          }
        }
      }
    }
    
  3. Restart OpenClaw.

Configuration

Option Type Default Description
dryRun boolean false Log what would be scrubbed without actually redacting
allowedRecipients string[] [] Chat IDs where scrubbing is skipped (e.g., private DMs with yourself)

Example

Before scrubbing:

SSH into [email protected] and check the service on localhost:8096

After scrubbing:

SSH into [redacted-target] and check the service on [redacted-service]

安全使用建议
This implementation appears coherent and limited to local redaction. Before installing: (1) enable dryRun first to confirm what will be redacted and avoid accidental over-redaction, (2) review and if needed customize RULES to avoid false positives (e.g., usernames that look like hostnames), (3) limit allowedRecipients rather than leaving it empty if you want exemptions, (4) test in a non-production environment, and (5) note the package has no homepage/author metadata—install only if you trust the source or after reviewing the code (index.ts is short and readable).
功能分析
Type: OpenClaw Skill Name: content-scrubber Version: 1.0.0 The content-scrubber skill is a security-focused utility designed to prevent the accidental leakage of private infrastructure details (such as RFC 1918 IP addresses, localhost ports, and SSH targets) in outgoing messages. The implementation in index.ts uses deterministic regex-based redaction and operates as a message interceptor, aligning perfectly with its stated purpose without any evidence of malicious intent, data exfiltration, or unauthorized execution.
能力评估
Purpose & Capability
Name, description, SKILL.md, and index.ts all implement the same capability: intercept outgoing messages and apply regex rules to redact IPs, localhost ports, hostnames, and SSH targets. No unrelated capabilities (cloud access, filesystem scanning, network exfiltration) are requested or implemented.
Instruction Scope
SKILL.md installation and config steps match the plugin code behavior (registers a message_sending hook, supports dryRun and allowedRecipients). The runtime instructions do not ask the agent to read unrelated files, environment variables, or transmit data to external endpoints.
Install Mechanism
No install spec or remote downloads are used; the skill is distributed as source files to be copied into the extensions directory. This is low-risk and proportional to the stated purpose.
Credentials
The skill requires no environment variables, credentials, or config paths. The plugin reads only its provided configuration (dryRun, allowedRecipients) and runtime event/context fields, which is appropriate.
Persistence & Privilege
always is false (default) and the plugin runs only when enabled by adding it to openclaw.json. It does not modify other plugins or global agent settings.
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 content-scrubber
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /content-scrubber 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 Content Scrubber for OpenClaw: - Redacts RFC 1918 IPs, loopback addresses, localhost ports, SSH/SCP targets, and custom hostnames from outgoing messages. - Uses fast and auditable regex-based rules, not LLM methods. - Configurable via `dryRun` mode and `allowedRecipients` for fine-grained control. - Integrates as a message interceptor for Discord, Telegram, Signal, and other messaging platforms. - Easy installation and setup in the OpenClaw extensions directory.
元数据
Slug content-scrubber
版本 1.0.0
许可证 MIT-0
累计安装 0
当前安装数 0
历史版本数 1
常见问题

Content Scrubber 是什么?

OpenClaw plugin that scrubs private infrastructure details from outgoing messages. Regex-based redaction of RFC 1918 IPs, localhost ports, SSH targets, and h... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 150 次。

如何安装 Content Scrubber?

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

Content Scrubber 是免费的吗?

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

Content Scrubber 支持哪些平台?

Content Scrubber 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 Content Scrubber?

由 Solomon Neas(@solomonneas)开发并维护,当前版本 v1.0.0。

💬 留言讨论