Skip to main content

Add your description here

Project description

PageEyes Agent

PageEyes Agent 是基于 Pydantic AI 框架开发的一个轻量级 UI Agent, 其中元素信息感知能力依靠 OmniParserV2 模型,整个 Agent 的优势在于不依赖视觉语言大模型, 即使小参数的 LLM 也能胜任路径规划能力,同时支持多平台(Web、Android),目前主要包含以下功能:

  1. 完全由自然语言指令驱动,无需编写脚本,既可实现自动化测试,UI巡检等任务
  2. 跨平台、夸端支持,在 Python 环境中安装 page-eyes 库和配置 OmniParser 服务后即可开始 Web、Android 平台的自动化任务,未来还将继续支持iOS平台
  3. 支持多种大模型接入,包括DeepSeek、OpenAI、千问等,默认使用 DeepSeek V3 模型,后续会支持更多大模型接入
  4. 可通过自然语言进行断言,并生成详细的执行日志和报告,方便测试人员查看执行过程和结果

安装

pip install page-eyes

使用示例

import asyncio

from page_eyes.agent import WebAgent, MobileAgent


async def main():
    # Web 端
    ui_agent = await WebAgent.create(simulate_device='iPhone 15 Pro')

    # 移动端
    # ui_agent = await MobileAgent.create(serial='android-udid')

    report = await ui_agent.run(
        ('1.打开 url "https://yobang.tencentmusic.com/chart/uni-chart/rankList/"\n'
         '2.点击"查找icon"\n'
         '3.在搜索输入框中输入"小美满"\n'
         '4.点击"小美满> "\n'
         '5.点击"日榜"'
         ))


if __name__ == "__main__":
    asyncio.run(main())

更新日志

  • 2025-09-03

    1. 支持等待、停留等自然语言操作
    2. 支持传递 OmniParser Key 鉴权
    3. 断言支持判断页面/屏幕中是否存在某个关键字
  • 2025-08-01

    1. 精简了系统提示词,每次调用使用更少的token
    2. 优化报告的记录和生成逻辑,解决步骤偶然错乱的问题
    3. 优化工具的参数,实现最小化参数,减少token
    4. 优化了等待逻辑,实现更精准的等待,不需要统一等待1s
    5. 截图、解析改成异步,并发不阻塞
    6. 优化浏览器配置,支持使用持久化缓存,二次启动页面速度更快
  • 2025-07-28

    1. 增加滚动和滑动操作
    2. 支持H5 swiper 组件滑动
  • 2025-06-25

    1. 优化 LLM Graph 控制
    2. 更新配置模式
  • 2025-06-09

    1. 支持 Web Agent 模式
    2. 增加执行报告和过程实时渲染
    3. 重构同步方法为异步
  • 2025-06-04

    1. 支持移动端 Agent 模式
    2. 增加click input open_url 工具
    3. 支持接入 OmniParser 进行元素解析

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

page_eyes-0.2.0.tar.gz (236.7 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

page_eyes-0.2.0-py3-none-any.whl (182.4 kB view details)

Uploaded Python 3

File details

Details for the file page_eyes-0.2.0.tar.gz.

File metadata

  • Download URL: page_eyes-0.2.0.tar.gz
  • Upload date:
  • Size: 236.7 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: uv/0.6.9

File hashes

Hashes for page_eyes-0.2.0.tar.gz
Algorithm Hash digest
SHA256 479ae1ec628ebc70dfff776e70da323a457015d1b86a1149919036238a832a14
MD5 b3128f57b5ad2953f3229e34fac83d15
BLAKE2b-256 8e2438e6d153eae3c7574b283802c06065fbde5fd22dc5c0b5a2a568af1c289d

See more details on using hashes here.

File details

Details for the file page_eyes-0.2.0-py3-none-any.whl.

File metadata

  • Download URL: page_eyes-0.2.0-py3-none-any.whl
  • Upload date:
  • Size: 182.4 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: uv/0.6.9

File hashes

Hashes for page_eyes-0.2.0-py3-none-any.whl
Algorithm Hash digest
SHA256 f1b3203404ac3590003bc390b553bd8facf0e4cfaa3c667c53af97f076830bd7
MD5 36a74f8f791ad37aaef4dc4e116b210b
BLAKE2b-256 5d1421c784e37a4ba03784fb461d6a4e85a01e3dceff5966b0c26a17dfdecd5f

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page