← 返回 Skills 市场
Rotate Captcha Solver
作者
yuxiaowu3000
· GitHub ↗
· v1.0.0
· MIT-0
83
总下载
0
收藏
0
当前安装
1
版本数
在 OpenClaw 中安装
/install rotate-captcha-solver
功能描述
专门处理旋转验证码的浏览器自动化技能。使用 OpenCV 边缘检测 + 霍夫变换识别图片倾斜角度,模拟人类拖动轨迹完成旋转验证。支持百度地图、各类需要"旋转图片到正确角度"的验证码。当遇到"旋转验证码"、"拖动滑块旋转"时自动触发。
安全使用建议
This skill appears to implement what it says (detect angle with OpenCV and simulate a slider drag), but be aware of several practical and security-relevant issues before installing:
- Missing dependency/install steps: The scripts require python3, opencv-python, numpy, and a Playwright-compatible agent-browser environment; examples use jq. Add explicit install instructions or install these dependencies in a controlled environment before running.
- Instruction/implementation mismatch: The recognizer outputs numeric coordinates in JSON, but the execute_rotation.js code expects a Playwright element handle and calls boundingBox() on it; findSliderElement() does not use the passed coordinates. Expect to modify one side (either find element by coordinates or change recognizer to return a selector/element reference) to make it work.
- Fragile internals: execute_rotation.js accesses internal Playwright internals (slider._context._page) and relies on specific DOM selectors — this can break across Playwright versions or site implementations.
- Ethical/legal: The skill automates bypassing visual verification controls (CAPTCHAs). Using it may violate websites' terms of service and could be used for abusive automation. Only use in authorized/testing contexts.
- Safety suggestions: Run the skill in a sandboxed agent environment, verify and pin dependency versions, add explicit install and permission notes to SKILL.md, and test on non-production/example pages first. If you plan to use it in production, fix the selector/coordinate mismatch and prefer robust methods for mapping coordinates -> page element bounding boxes.
功能分析
Type: OpenClaw Skill
Name: rotate-captcha-solver
Version: 1.0.0
The skill bundle is a legitimate tool for automating the solving of rotation-based captchas using OpenCV and Playwright. The Python script (recognize_rotation.py) uses standard image processing techniques like Canny edge detection and Hough transforms, while the JavaScript component (execute_rotation.js) implements human-like mouse trajectories to avoid bot detection. No evidence of data exfiltration, malicious execution, or prompt injection was found; all behaviors align with the stated purpose of browser automation for captcha resolution.
能力评估
Purpose & Capability
The skill's name/description align with the provided Python (OpenCV) and JavaScript (Playwright) scripts which implement detection and simulated dragging. However the SKILL.md does not declare required runtime dependencies (python3, opencv-python, numpy, jq, Playwright/agent-browser), nor does it provide install steps — this omission reduces coherence about how to actually run the skill.
Instruction Scope
SKILL.md instructs taking screenshots, running the Python recognizer to produce a JSON, then calling executeRotation in the browser. But there's a logical mismatch: recognize_rotation.py outputs numeric coordinates (slider_location) while execute_rotation.js expects a Playwright element handle (calls slider.boundingBox()). The provided findSliderElement() ignores the passed location and only searches DOM selectors, with no fallback to using coordinates. The examples also invoke shell helpers like jq without declaring them. These inconsistencies mean the runtime instructions may fail or require code changes.
Install Mechanism
No install spec is provided (instruction-only), which minimizes installer risk. But the code depends on third-party libraries (OpenCV, numpy, Playwright) and CLI tools (jq) with no guidance; missing dependency installation is an operational risk rather than a supply-chain red flag.
Credentials
The skill does not request environment variables, credentials, or config paths and the code does not access external credentials or network endpoints. All file I/O is local (screenshots, debug images, result JSON).
Persistence & Privilege
The skill is not always-enabled and has no special persistence or privilege changes. It does not modify other skills or system 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 rotate-captcha-solver - 安装完成后,直接呼叫该 Skill 的名称或使用
/rotate-captcha-solver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
Initial release of rotate-captcha-solver: a browser automation skill for solving rotate image captchas.
- Detects and solves "rotate image to correct angle" captchas (e.g., Baidu Maps, login/register pages).
- Uses OpenCV edge detection and Hough Transform to calculate skew angle.
- Simulates human-like slider drag trajectories for rotation.
- Includes scripts for rotation angle recognition and rotation execution.
- Supports debug modes, batch processing, and trajectory customization.
- Provides troubleshooting guidance and FAQ for improved captcha solve rates.
元数据
常见问题
Rotate Captcha Solver 是什么?
专门处理旋转验证码的浏览器自动化技能。使用 OpenCV 边缘检测 + 霍夫变换识别图片倾斜角度,模拟人类拖动轨迹完成旋转验证。支持百度地图、各类需要"旋转图片到正确角度"的验证码。当遇到"旋转验证码"、"拖动滑块旋转"时自动触发。 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 83 次。
如何安装 Rotate Captcha Solver?
在 OpenClaw 或 Claude Code 对话框中运行命令「/install rotate-captcha-solver」即可一键安装,无需额外配置。
Rotate Captcha Solver 是免费的吗?
是的,Rotate Captcha Solver 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。
Rotate Captcha Solver 支持哪些平台?
Rotate Captcha Solver 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 Rotate Captcha Solver?
由 yuxiaowu3000(@yuxiaowu3000)开发并维护,当前版本 v1.0.0。
推荐 Skills