OCR Space: Free OCR API

调用 OCR.space 免费 API 识别图片中的文字

Security Scan

The skill's code and instructions match its stated purpose (calling OCR.space with the free test key); nothing requests unrelated credentials or installs arbitrary code, though it does rely on Python packages that are not declared and it uploads user images to a third-party API.

Purpose & Capability

Name/description match implementation: the script posts images to https://api.ocr.space/parse/image using the free test key ('helloworld') and returns parsed text. No unrelated services, binaries, or credentials are requested.

Instruction Scope

SKILL.md and the script confine actions to reading the provided image path (or accepting base64), optionally compressing it, and POSTing it to the OCR.space API. There are no instructions to read other system files, environment secrets, or to exfiltrate data to unexpected endpoints.

Install Mechanism

No install spec (instruction-only) and no network downloads, which is low risk. However the included Python script requires third-party packages ('requests' and Pillow) that are not declared in the metadata or SKILL.md; the skill will fail or raise import errors unless those are installed.

Credentials

No environment variables or credentials are required. The script hardcodes API_KEY='helloworld' (the documented free test key) which matches the SKILL.md. If a user wants to use a private paid key they would need to change the script or add a mechanism to supply it.

Persistence & Privilege

Skill is not always-enabled, does not request persistent platform-wide privileges, and does not modify other skills or global agent settings.

Assessment

This skill appears to do what it says: it reads images and uploads them to OCR.space using the free test key. Before installing, consider: (1) it will transmit image data to a third-party server — don't send sensitive images unless you accept that. (2) You need Python packages 'requests' and 'Pillow' installed; SKILL.md doesn't list them, so install them (pip install requests pillow) or run the script in a controlled environment. (3) If you plan to use a paid API key, remove the hardcoded 'helloworld' and provide your key securely (avoid committing it into the file). (4) Test the script locally to confirm behavior and network calls before allowing any agent to invoke it autonomously.

OCR.space 文字识别 Skill

使用免费 OCR API 识别图片中的文字，支持多语言。

🚀 功能

• ✅ 图片文字识别 (OCR)
• ✅ 多语言支持
• ✅ 免费无需注册
• ✅ 支持中文、英文、日文、韩文等 20+ 语言

---

📡 OCR.space 免费版限制

限制项	额度
每日请求数	500 次/天
单文件大小	5 MB（自动压缩）
API Key	不需要（测试用 helloworld）

⚠️ 注意：免费版适合测试和小规模使用，生产环境建议购买付费版

---

🌍 支持的语言

语言代码	语言名称
chs	中文简体
cht	中文繁体
eng	英语
jpn	日语
kor	韩语
fre	法语
ger	德语
spa	西班牙语
por	葡萄牙语
rus	俄语
ara	阿拉伯语
tha	泰语
hin	印地语
vie	越南语
ita	意大利语
dut	荷兰语
pol	波兰语
tur	土耳其语

💡 Engine 2 和 Engine 3 支持语言自动检测

---

📁 文件结构

ocr-space/
├── SKILL.md          # 本文件
├── ocr_space.py      # 主脚本
└── README.md         # 使用说明

---

🔧 使用方法

命令行

# 识别中文（默认）
python3 ocr_space.py /path/to/image.jpg

# 识别英文
python3 ocr_space.py /path/to/image.jpg eng

# 识别日文
python3 ocr_space.py /path/to/image.png jpn

Python 代码调用

import sys
sys.path.insert(0, '/Users/shusiwei/.openclaw/workspace/skills/ocr-space')
from ocr_space import ocr_image

# 识别图片
result = ocr_image('/path/to/image.jpg', language='chs')
if result:
    print(result)

---

📝 OpenClaw 中使用

在 OpenClaw 中可以通过 exec 调用：

python3 ~/.openclaw/workspace/skills/ocr-space/ocr_space.py <图片路径> [语言]

示例：

# 识别桌面图片
python3 ~/.openclaw/workspace/skills/ocr-space/ocr_space.py ~/Desktop/screenshot.png

---

🔍 完整 API 文档

更多参数和使用方法请参考：OCR.space OCR API

---

📜 更新日志

• 2026-03-10: 初始版本，支持基本 OCR 功能