← 返回 Skills 市场
屏幕截图OCR工具
作者
flyingants
· GitHub ↗
· v1.0.1
· MIT-0
256
总下载
0
收藏
0
当前安装
2
版本数
在 OpenClaw 中安装
/install screenshot-ocr-tool
功能描述
从屏幕截图中提取文字，支持纯文本、结构化及问答分离格式，适合中医考试题识别。
安全使用建议
This skill appears coherent and local-only, but before installing: 1) ensure you have Tesseract OCR installed (and Chinese traineddata 'chi_sim') and Python packages Pillow and pytesseract — the skill does not declare or install them; 2) test it in a safe/isolated environment to confirm it behaves as expected; 3) be aware that image contents are processed locally by the script — do not feed sensitive images unless you are comfortable with local processing; 4) note minor metadata issues (SKILL.json version vs registry, duplicate function in script) that suggest light maintenance issues but not malicious intent. If you need automatic dependency installation or stronger provenance (homepage/source), request those from the author before trusting in production.
功能分析
Type: OpenClaw Skill
Name: screenshot-ocr-tool
Version: 1.0.1
The skill bundle is a legitimate OCR tool designed for extracting and structuring text from screenshots, specifically tailored for exam questions. The implementation in `scripts/screenshot_ocr.py` uses standard libraries like `pytesseract` and `PIL` for image processing and regular expressions for text parsing. No evidence of data exfiltration, malicious execution, or prompt injection was found.
能力评估
Purpose & Capability
Code implements OCR + structure/QA parsing consistent with the skill description (TCM exam screenshots). However, the skill uses Python packages (pytesseract, Pillow) and the Tesseract engine (system binary and traineddata 'chi_sim') but does not declare these runtime requirements in the metadata — this is an alignment/packaging omission.
Instruction Scope
SKILL.md points the agent to ./scripts/screenshot_ocr.py and the script only opens the provided image path, runs local OCR, and parses text. The instructions and implementation do not access unrelated files, environment variables, or network endpoints.
Install Mechanism
No install spec (instruction-only) — lowest installation risk. Code files are included but no automated installer. Because the code depends on Python packages and the external tesseract binary, the skill will fail without those installed; the skill does not provide an installation step or list these dependencies.
Credentials
The skill requests no environment variables, no credentials, and references no config paths. This is proportionate to its stated purpose (local OCR).
Persistence & Privilege
always is false and the skill does not request persistent or elevated agent privileges. It does not modify other skills or system-wide 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 screenshot-ocr-tool - 安装完成后，直接呼叫该 Skill 的名称或使用
/screenshot-ocr-tool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
v1.0.1 - Bump version
v1.0.0
v1.0.0 - First release
元数据
常见问题
屏幕截图OCR工具 是什么？
从屏幕截图中提取文字，支持纯文本、结构化及问答分离格式，适合中医考试题识别。 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 256 次。
如何安装 屏幕截图OCR工具？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install screenshot-ocr-tool」即可一键安装，无需额外配置。
屏幕截图OCR工具 是免费的吗？
是的，屏幕截图OCR工具 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
屏幕截图OCR工具 支持哪些平台？
屏幕截图OCR工具 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 屏幕截图OCR工具？
由 flyingants（@tom859174-sketch）开发并维护，当前版本 v1.0.1。
推荐 Skills