← 返回 Skills 市场
325
总下载
0
收藏
0
当前安装
16
版本数
在 OpenClaw 中安装
/install shi-gou
功能描述
监控权限拒绝事件，检测提示词注入、路径遍历及危险命令，生成安全报告并识别潜在威胁告警。
安全使用建议
This skill appears coherent and focused on local pattern-based scanning and sanitization. Before installing, confirm the agent runtime supports executing the included JavaScript (Node.js or equivalent), since the README references Node.js but the registry metadata does not declare any required binaries. Also verify the platform will not automatically grant network access or secret-scope elevation to the skill at runtime (the provided code makes no network calls or credential reads). If you require integration with other systems (logs, databases, or external alerting), expect to extend the skill and re-evaluate required permissions at that time.
功能分析
Type: OpenClaw Skill
Name: shi-gou
Version: 2.0.6
The 'shi-gou' (Security Sentinel) skill is a security utility designed to detect prompt injection, dangerous commands, and path traversal patterns. The core logic in `scripts/security-check.js` uses standard regular expressions to identify threats and redact sensitive information like API keys and internal IP addresses, showing no evidence of malicious intent, data exfiltration, or unauthorized execution.
能力标签
能力评估
Purpose & Capability
The name, description, SKILL.md, README, and scripts/security-check.js all align: the skill performs prompt-injection, path-traversal, and dangerous-command detection plus sanitization and reporting. Minor inconsistency: README mentions Node.js as the runtime but the registry metadata lists no required binaries; the code is JavaScript, so a runtime capable of executing the script (Node.js or equivalent) is implied but not declared.
Instruction Scope
SKILL.md keeps scope to scanning inputs, generating reports, and sanitizing commands. The instructions do not ask the agent to read unrelated files, environment variables, or send data to external endpoints. The included script likewise only analyzes strings and returns simulated reports (no I/O or network).
Install Mechanism
No install spec is provided (instruction-only metadata), and the included JS file has no external downloads or install steps. This is low-risk; the skill will rely on the agent runtime to execute the provided script.
Credentials
The skill declares no required environment variables, credentials, or config paths. The code does not access environment variables or secrets. The lack of requested credentials is proportionate to its local-analysis functionality.
Persistence & Privilege
always is false and autonomous invocation is allowed (platform default). The skill does not request system-wide configuration changes or long-term privileges; no elevated persistence or cross-skill config modification is present.
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 shi-gou - 安装完成后，直接呼叫该 Skill 的名称或使用
/shi-gou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
v2.0.6
- 更新版本号至 2.0.6，并同步所有 changelog 项及历史版本号
- 无功能/能力变更，仅为 changelog 和元信息修订，保持与实际版本号一致
- 技能能力与接口保持不变
v2.0.5
shi-gou 2.0.5
- 更新版本号与变更历史（changelog）中的所有版本号为 2.0.5
- 其余内容未发生更改
v2.0.4
- Updated changelog version numbers from 2.0.3 to 2.0.4 for all historical entries.
- No changes to features or functionality; documentation version alignment only.
v2.0.3
- Changelog version numbers updated from 2.0.2 to 2.0.3 to match the current release.
- No changes to features or functionality; only historical version numbers were revised for accuracy in the changelog section.
v2.0.2
- Updated changelog and metadata to version 2.0.2.
- No code or functional changes; only the SKILL.md file was modified.
- The version history and changelog entries were revised to reflect the new version number.
v2.0.1
- Migrated to v2 skill specification format for enhanced structure and metadata.
- Added progressive/lazy loading and capacity management settings.
- Standardized soul attribution and added more detailed trigger keywords.
- Expanded documentation for capabilities and usage scenarios.
v1.0.11
Version 1.0.11
- Documentation updates in SKILL.md; no logic or API changes.
- Updated version header to 1.0.11.
- No functional modifications to skill capabilities.
v1.0.10
- Update version to 1.0.10 in SKILL.md.
- No functional or content changes; only the version header updated.
v1.0.9
shi-gou 1.0.9
- 更新 SKILL.md 版本号至 1.0.9
- 其他内容未变
v1.0.8
- Bumped version to 1.0.8 in SKILL.md.
- No other content changes detected.
v1.0.7
- 更新 version 字段至 1.0.7
- 其余内容未变，未引入新功能或接口修改
v1.0.6
- Bumped version to 1.0.6 in SKILL.md.
- No other content or functionality changes.
v1.0.5
- Updated version to 1.0.5 in documentation.
- No other content or feature changes.
v1.0.4
- Bump version to 1.0.4 in SKILL.md.
- No other content or functionality changes.
v1.0.3
shi-gou 1.0.3
- Updated SKILL.md version to 1.0.3.
- No functional or capability changes—documentation update only.
v1.0.2
shi-gou v1.0.2
- Clarified core responsibilities and capabilities in SKILL.md, including specific threat detection patterns and outputs.
- Added detailed parameter and output specifications for each function (security_check, security_report, sanitize_command).
- Improved documentation on integration with other skills and the overall seven-po architecture.
- No code/configuration changes noted; focus is on clearer skill structure and usage instructions.
元数据
常见问题
Shi Gou 是什么？
监控权限拒绝事件，检测提示词注入、路径遍历及危险命令，生成安全报告并识别潜在威胁告警。 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 325 次。
如何安装 Shi Gou？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install shi-gou」即可一键安装，无需额外配置。
Shi Gou 是免费的吗？
是的，Shi Gou 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
Shi Gou 支持哪些平台？
Shi Gou 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 Shi Gou？
由 无为上人（@lt8899789）开发并维护，当前版本 v2.0.6。
推荐 Skills