← 返回 Skills 市场
Camsnap
作者
Peter Steinberger
· GitHub ↗
· v1.0.0
18874
总下载
13
收藏
1154
当前安装
1
版本数
在 OpenClaw 中安装
/install camsnap
功能描述
Capture frames or clips from RTSP/ONVIF cameras.
安全使用建议
Install only if you trust the camsnap CLI and are comfortable giving the agent access to your camera endpoints. Store camera credentials carefully, avoid broad or destructive `--action` commands, and require explicit user intent before starting watch mode or writing captured media to shared locations.
功能分析
Type: OpenClaw Skill
Name: camsnap
Version: 1.0.0
The skill bundle is classified as suspicious due to instructions for the agent to handle sensitive credentials (username/password) directly in `SKILL.md` for camera configuration. Additionally, the `camsnap watch` command includes an `--action` parameter, which implies the agent could be instructed to execute arbitrary commands, presenting a potential for malicious execution if misused. While these capabilities are plausibly needed for the stated purpose of camera interaction, they represent high-risk operations for an AI agent.
能力评估
Purpose & Capability
The skill purpose is coherent: it documents using camsnap to capture snapshots, clips, and motion events from RTSP/ONVIF cameras. ([clawhub.ai](https://clawhub.ai/steipete/camsnap))
Instruction Scope
The documented `camsnap watch ... --action '...'` hook is powerful, but it is disclosed as part of motion-event automation and there is no artifact evidence of hidden or automatic command execution. ([clawhub.ai](https://clawhub.ai/steipete/camsnap))
Install Mechanism
The install surface is a normal OpenClaw skill install, with the skill page showing command documentation rather than bundled executable code; it also notes an external `ffmpeg` requirement. ([clawhub.ai](https://clawhub.ai/steipete/camsnap))
Credentials
Camera hostnames, usernames, passwords, local config, and media outputs are sensitive, but they are expected for managing RTSP/ONVIF cameras and proportional to the stated purpose. ([clawhub.ai](https://clawhub.ai/steipete/camsnap))
Persistence & Privilege
Watch mode can be long-running and can trigger an action hook, but I found no evidence of stealth persistence, privilege escalation, exfiltration, or destructive behavior.
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 camsnap - 安装完成后，直接呼叫该 Skill 的名称或使用
/camsnap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
v1.0.0
元数据
常见问题
Camsnap 是什么？
Capture frames or clips from RTSP/ONVIF cameras. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 18874 次。
如何安装 Camsnap？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install camsnap」即可一键安装，无需额外配置。
Camsnap 是免费的吗？
是的，Camsnap 完全免费（开源免费），可自由下载、安装和使用。
Camsnap 支持哪些平台？
Camsnap 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 Camsnap？
由 Peter Steinberger（@steipete）开发并维护，当前版本 v1.0.0。
推荐 Skills