← 返回 Skills 市场
47
总下载
0
收藏
0
当前安装
2
版本数
在 OpenClaw 中安装
/install prod-shield
功能描述
ProdShield: Hardened execution guardrails — because production only gets one chance. Use this skill whenever Claude is about to execute commands, run scripts...
安全使用建议
This skill appears to be what it claims: a conservative, instruction-only guardrail for destructive operations. Before installing, check the missing pieces: (1) confirm the exact confirmation phrase and how the agent will match it (Section 6 was truncated in the provided copy), (2) test the skill in a safe environment to measure false positives and ensure it doesn't block legitimate automation, and (3) verify how it integrates with your agent pipeline — make sure other skills and automation flows will actually call or respect ProdShield when needed. Because it defaults ambiguous targets to 'production', expect extra prompts; if that's too noisy for your workflows, consider customizing the environment patterns or whitelist rules. Finally, since this is instruction-only, review the full SKILL.md (complete Section 6) and the confirmation/confirmation-handling behavior before relying on it in critical automation.
功能分析
Type: OpenClaw Skill
Name: prod-shield
Version: 1.0.1
The 'prod-shield' skill is a safety-oriented instruction bundle designed to prevent accidental destruction of production environments by an AI agent. It implements comprehensive guardrails, including environment detection (SKILL.md), a catalog of dangerous commands (references/dangerous-commands.md), and strict confirmation protocols for destructive actions. The skill contains no executable code, external dependencies, or instructions for data exfiltration, and its logic is entirely aligned with its stated purpose of system protection.
能力评估
Purpose & Capability
Name/description claim a production safety guardrail and the skill only includes instruction files describing environment detection, dangerous-command patterns, and recovery playbooks. It requests no credentials, binaries, installs, or config paths — all proportionate to a safety-only skill.
Instruction Scope
SKILL.md prescribes strict detection rules, pre-execution checklists, prohibited-actions lists, dry-run and confirmation protocols, and recovery playbooks — all within the scope of preventing destructive operations. The guidance is conservative (defaults ambiguous targets to production), which is intentional but may produce many false positives or block legitimate automated workflows. The distributed docs are truncated where Section 6 confirmation phrasing would appear; you should verify the exact confirmation phrase and how the skill expects confirmation to be provided.
Install Mechanism
Instruction-only skill with no install steps, no downloads, and no code to execute — minimal installation risk.
Credentials
The skill requests no environment variables, secrets, or external credentials. It only reads contextual names/URLs/connection strings from the immediate prompt context (as required to decide whether a target is production), which aligns with its stated purpose.
Persistence & Privilege
The skill is not always-enabled and is user-invocable; it does not request elevated permanence or modify other skills. Autonomous invocation is enabled by default on the platform, which is normal; nothing in the package attempts to force permanent inclusion or to change other skill configurations.
如何使用
- 确保已安装 OpenClaw（本地或 Docker 部署）
- 在对话框中输入安装命令：
/install prod-shield - 安装完成后，直接呼叫该 Skill 的名称或使用
/prod-shield触发 - 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
- Clarified that the skill is purely instruction-only and requires no scripts, binaries, or install steps.
- Changed metadata: removed "always: true" and added "userInvocable: true" for improved invocation flexibility.
- Updated compatibility section to stress "drop the folder and it works immediately" on all OpenClaw setups.
- Marked dependencies as "instruction-only, zero external binaries required" for added clarity.
- Minor formatting and wording changes for better readability and user guidance.
v1.0.0
ProdShield is now fully hardened for every OpenClaw version and platform. Here's what changed to ensure universal compatibility:
always: true in metadata — ProdShield loads unconditionally on every OpenClaw version, regardless of environment, config gating, or binary checks. It never gets filtered out.
Zero dependencies — No requires.bins, no requires.env, no external tools needed. It works on npm installs, git installs, Docker, Railway, Render, Hetzner, Fly.io — every deployment method.
Cross-platform dangerous command coverage — Added Windows-native equivalents alongside Unix commands: Remove-Item -Recurse -Force, rd /s /q, del /f /s /q, Windows temp paths (%TEMP%, AppData\Local\Temp), and PowerShell destructive patterns.
All agent modes covered — Single-agent, multi-agent, sandboxed, and elevated. Since it's pure instructions (no binary execution), it works identically in sandboxed containers too.
All channels — Works over WhatsApp, Telegram, Discord, Slack, iMessage — wherever OpenClaw runs, ProdShield runs with it.
元数据
常见问题
Hardened execution guardrails — because production only gets one chance. 是什么？
ProdShield: Hardened execution guardrails — because production only gets one chance. Use this skill whenever Claude is about to execute commands, run scripts... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 47 次。
如何安装 Hardened execution guardrails — because production only gets one chance.？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install prod-shield」即可一键安装，无需额外配置。
Hardened execution guardrails — because production only gets one chance. 是免费的吗？
是的，Hardened execution guardrails — because production only gets one chance. 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
Hardened execution guardrails — because production only gets one chance. 支持哪些平台？
Hardened execution guardrails — because production only gets one chance. 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 Hardened execution guardrails — because production only gets one chance.？
由 2eux（@2eux）开发并维护，当前版本 v1.0.1。
推荐 Skills