← 返回 Skills 市场
openclawselfguard
作者
aaronjager92
· GitHub ↗
· v1.0.1
· MIT-0
104
总下载
0
收藏
0
当前安装
2
版本数
在 OpenClaw 中安装
/install openclaw-self-guard
功能描述
Monitors local OpenClaw version daily at 06:00 Beijing time against NVD and GitHub advisories, reporting found CVEs with remediation steps.
安全使用建议
This skill appears to do what it says: detect the local OpenClaw version, query NVD and GitHub, and schedule daily checks. Before installing or running it: 1) Inspect and (if desired) run the scripts manually to verify behavior (python3 scripts/get_version.py --json and python3 scripts/check_vulns.py --json). 2) Review and approve the cron/job entry that setup_cron.sh will write to ~/.openclaw/cron/jobs.json (the script backs up existing file). 3) Do not pass a delivery channel to setup_cron.sh unless you trust the configured channel — by default it is console only; enabling a channel may cause the report (including local version info and any findings) to be sent externally. 4) Be aware of bugs: fetch_github.py has a coding error (an undefined 'undefined' usage in params) and version-range checking in check_vulns.py is simplistic; these may make the tool fail or produce false negatives/positives. If you want to use it, consider fixing those issues or running checks manually and verifying outputs before relying on the cron automation.
功能分析
Type: OpenClaw Skill
Name: openclaw-self-guard
Version: 1.0.1
The skill is a security monitoring tool that checks for OpenClaw vulnerabilities using NVD and GitHub APIs. It performs high-risk actions including automated persistence by modifying the OpenClaw cron configuration (~/.openclaw/cron/jobs.json), executing local shell commands for version detection (openclaw --version), and making external network requests to services.nvd.nist.gov and api.github.com. While these capabilities are aligned with the stated purpose, the combination of automated system modification and logic flaws—such as a broken version comparison in check_vulns.py and a NameError (undefined variable) in fetch_github.py—warrants a suspicious classification under the provided criteria.
能力评估
Purpose & Capability
Name/description state: daily checks of local OpenClaw against NVD and GitHub advisories. The repository includes scripts to detect local version, query NVD and GitHub, compare versions, format reports, and install a cron job — all consistent with the stated purpose.
Instruction Scope
SKILL.md and setup_cron.sh instruct the agent to auto-install a daily job that runs the supplied check_vulns.py script. The cron payload explicitly tells the agent to run python3 check_vulns.py --json and return either a full report (if vulnerabilities found) or a short success message. By default the delivery in the job is 'console only', but the job file supports configuring external delivery channels (feishu, telegram, etc.). This means the output could be sent externally if the user or the setup script is run with a channel argument — review jobs.json and chosen channel before enabling.
Install Mechanism
No remote install spec (instruction-only) — low risk for remote code fetch. However setup_cron.sh modifies user config under ~/.openclaw/cron/jobs.json and creates backups; installing the skill will therefore persist a scheduled job in the user's home config. That persistent change is expected for a monitoring skill but is a meaningful modification to user config and should be reviewed.
Credentials
The skill requests no environment variables or credentials. Network calls go to known public sources (nist.gov and api.github.com). The scripts read some local files/paths to detect OpenClaw (openclaw --version and a few package.json locations) which is appropriate for detecting a local version.
Persistence & Privilege
The skill installs a persistent daily job (cron-like) in ~/.openclaw which will cause repeated autonomous checks. always:false (not force-included). Autonomous invocation via the agent/job system is expected for this monitoring use case, but persistent scheduled runs increase the blast radius if the delivery channel is changed to an external recipient — verify delivery settings.
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 openclaw-self-guard - 安装完成后,直接呼叫该 Skill 的名称或使用
/openclaw-self-guard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 and simplified data sources: now only NVD and GitHub Security Advisories are listed.
- Updated cron job section: clarified that delivery is console output by default, with instructions for customizing the output channel.
- Removed references to international/intelligence sources and CVE.org for improved accuracy.
- Updated features and usage documentation for clarity and conciseness.
- Updated requirements to include `lxml` in addition to `requests` and `beautifulsoup4`.
v1.0.0
Initial release of openclaw-self-guard – an automated security vulnerability monitor for OpenClaw.
- Automatically checks OpenClaw for CVEs via NVD, GitHub advisories, and international sources.
- Compares local version against known vulnerabilities and outputs detailed remediation steps if issues are detected.
- Runs silently if no vulnerabilities are found.
- Installs a daily cron job on setup (runs at 06:00 Beijing time).
- Provides manual check commands and clear output/report formats.
元数据
常见问题
openclawselfguard 是什么?
Monitors local OpenClaw version daily at 06:00 Beijing time against NVD and GitHub advisories, reporting found CVEs with remediation steps. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 104 次。
如何安装 openclawselfguard?
在 OpenClaw 或 Claude Code 对话框中运行命令「/install openclaw-self-guard」即可一键安装,无需额外配置。
openclawselfguard 是免费的吗?
是的,openclawselfguard 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。
openclawselfguard 支持哪些平台?
openclawselfguard 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 openclawselfguard?
由 aaronjager92(@aaronjager92)开发并维护,当前版本 v1.0.1。
推荐 Skills