Captcha Solver

验证码识别与解决 - 本地OCR识别 + 第三方API / CAPTCHA Recognition and Solving - Local OCR + Third-party APIs

Security Scan

The skill's code, instructions, and requirements are consistent with a CAPTCHA solving tool; it asks for no unrelated credentials or elevated privileges, but it sends images and API keys to third-party CAPTCHA services (over plain HTTP in the code) and can be used for abusive purposes — review before use.

Purpose & Capability

Name/description match the provided code and SKILL.md: the package implements local OCR, slide-gap detection, and optional third‑party API solving (2Captcha). There are no unrelated credentials or unexpected binaries requested.

Instruction Scope

SKILL.md and README show expected usage (local OCR, slide-solving, reCaptcha via APIs). The runtime instructions and examples align with the code. Note: the agent/code will read local image files, create a processed image in /tmp, and may send image data and site keys/URLs to third-party services — this is expected for the stated purpose but is a data-exfiltration vector for images/URLs provided to the tool.

Install Mechanism

No install spec (instruction-only) and the included Python script has normal library imports. No downloads from external/untrusted URLs or archive extraction are used by the skill itself.

Credentials

The metadata declares no required env vars, but the code reads API_2CAPTCHA (and also accepts an API key via CLI). Requesting a 2Captcha API key is proportional to the API-solving feature. However, the code uses plain HTTP endpoints (http://2captcha.com/...) which transmits API keys and base64 image payloads unencrypted — this is a security/privacy concern and should be changed to HTTPS.

Persistence & Privilege

The skill does not request persistent/system-wide presence, does not set always:true, and does not modify other skills or system configuration. It runs only when invoked.

Assessment

This skill is internally coherent for solving CAPTCHAs: it performs local OCR and optionally forwards CAPTCHAs to third‑party solving services (2Captcha). Before installing, consider: (1) legal/ethical risk — automated CAPTCHA solving can violate website terms or be used for abuse; (2) privacy risk — images you feed the tool will be sent to external services if you use API solving; (3) transport security — the included code uses unencrypted HTTP for 2Captcha calls (leaking API keys and image data); change to HTTPS (https://2captcha.com/...) or confirm TLS is used; (4) trust/code review — the script runs subprocesses (tesseract) and reads local files (/tmp), so run it in a controlled environment and inspect or pin the code; (5) if you will use third‑party APIs, provide a dedicated API key with limited funds/privileges and do not store high-value secrets in the environment. If you need higher assurance, request the author to switch API calls to HTTPS, document network endpoints explicitly in the SKILL.md, and list optional environment variables in the metadata.

验证码识别与解决 / CAPTCHA Solver

自动识别和解决各类验证码 / Automatically recognize and solve various CAPTCHAs

支持类型 / Supported Types

本地OCR识别 / Local OCR (免费/Free)

• 🔤 简单文本验证码 / Simple text CAPTCHA
• 🔢 数字验证码 / Numeric CAPTCHA
• ➕ 数学运算验证码 / Math CAPTCHA
• 🖼️ 滑动验证码(缺口检测) / Slide CAPTCHA (gap detection)

API解决 / API Solving (付费/APIs)

• reCAPTCHA v2/v3
• hCaptcha
• Cloudflare Turnstile
• 2Captcha / Anti-Captcha

使用方法 / Usage

# 识别图片验证码
python solve.py --image captcha.png

# 解决reCaptcha
python solve.py --recaptcha "site_key" --url "page_url"

# 滑动验证码
python solve.py --slide background.png --template slider.png

配置 / Configuration

本地OCR

# 默认使用Tesseract
TESSERACT_CMD = "/usr/bin/tesseract"
LANG = "eng+chi_sim"  # 支持中英文

API服务 (可选)

# 2Captcha
API_2CAPTCHA = "your_api_key"

# Anti-Captcha  
API_ANTI_CAPTCHA = "your_api_key"

算法 / Algorithms

1. 图像预处理

• 灰度转换 / Grayscale
• 二值化 / Binarization
• 去噪 / Denoising
• 锐化 / Sharpening

2. 字符分割

• 连通域分析 / Connected component analysis
• 投影法 / Projection method

3. 字符识别

• 模板匹配 / Template matching
• 机器学习 / ML-based OCR

4. 滑动验证码

• 边缘检测 / Edge detection
• 缺口定位 / Gap localization
• 轨迹生成 / Trajectory generation

示例 / Examples

简单文本识别

from solver import CaptchaSolver

solver = CaptchaSolver()
result = solver.solve_image("captcha.png")
print(result)  # 输出识别的字符

滑动验证码

result = solver.solve_slide(bg_img, slider_img)
print(result)  # 输出滑动距离

reCaptcha

result = solver.solve_recaptcha(site_key, page_url)
print(result)  # 输出token

服务对比 / Service Comparison

服务	价格	成功率	速度
本地OCR	免费	60-80%	快
2Captcha	$2.99/1000	95%+	慢
Anti-Captcha	$2.00/1000	95%+	中

注意事项 / Notes

1. 优先使用本地OCR，失败再调用API
2. 遵守网站使用条款
3. 不要用于非法用途