☰

TweetFilter AI

A highly customizable AI rates tweets 1-10 and removes all the slop, saving your braincells!

当前为 2025-04-15 提交的版本，查看 最新版本。

如何安装

您需要先安装一个扩展，例如 篡改猴、Greasemonkey 或 暴力猴，之后才能安装此脚本。

您需要先安装一个扩展，例如 篡改猴 或 暴力猴，之后才能安装此脚本。

您需要先安装一个扩展，例如 篡改猴 或 暴力猴，之后才能安装此脚本。

您需要先安装一个扩展，例如 篡改猴 或 Userscripts ，之后才能安装此脚本。

您需要先安装一款用户脚本管理器扩展，例如 Tampermonkey，才能安装此脚本。

您需要先安装用户脚本管理器扩展后才能安装此脚本。

（我已经安装了用户脚本管理器，让我安装！）

如何安装

您需要先安装一款用户样式管理器扩展，比如 Stylus，才能安装此样式。

您需要先安装一款用户样式管理器扩展，比如 Stylus，才能安装此样式。

您需要先安装一款用户样式管理器扩展，比如 Stylus，才能安装此样式。

您需要先安装一款用户样式管理器扩展后才能安装此样式。

您需要先安装一款用户样式管理器扩展后才能安装此样式。

您需要先安装一款用户样式管理器扩展后才能安装此样式。

（我已经安装了用户样式管理器，让我安装！）

作者
obsxrver
评分
0 0 0
版本
Version 1.3.6.3
创建于
2025-04-10
更新于
2025-04-15
大小
230.9 KB
许可证
MIT
适用于

X/Twitter De-Sloppifier

Version 1.3.6.3

Advanced tweet rating and filtering userscript with model selection, enhanced rating indicators, thread hierarchy mapping, and handle management.

Features

  • Tweet Rating: Automatically rates tweets on a scale from 1-10 using AI
  • Filtering System: Hide low-quality tweets based on your chosen threshold
  • Model Selection: Choose from multiple AI models for tweet rating
  • Image Processing: Dedicated image model for analyzing tweet images
  • Real-Time Response Streaming: Stream responses from the AI in real time
  • Thread Context: Improved thread hierarchy detection for better context awareness
  • Media Collection: Automatically collects all media from threads for comprehensive analysis
  • Handle Management: Auto-rate specific accounts as 10/10
  • Custom Instructions: Define your own criteria for how tweets should be rated
  • Dark Theme UI: Sleek interface matching X's design language
  • Parameter Control: Adjust temperature and top-p values for both models
Menu Display Model Selection
Tweet Analysis Example

Installation

  1. Install a userscript manager like Tampermonkey or Violentmonkey
  2. Install X/Twitter De-Sloppifier from GreasyFork
  3. Get an API key from OpenRouter
  4. Configure the script with your API key

Usage

After installation:

  1. Navigate to Twitter/X
  2. The minimum score filter will appear in the top-right corner
  3. Click "Open Menu" to access the full settings
  4. Input your OpenRouter API key
  5. Choose your preferred models and settings
  6. Save settings and enjoy a cleaner Twitter experience

Menu Display

Configuration

Rating Model Settings

  • Select any OpenRouter-compatible model for tweet rating
  • Adjust temperature and top-p parameters to control randomness
  • Create custom rating instructions to personalize scoring criteria

Image Model Settings

  • Choose a vision-capable model for image analysis
  • Control generation parameters separately from the main model

Handle Management

  • Add Twitter handles to automatically rate as 10/10
  • Useful for accounts you always want to see

How It Works

The userscript analyzes tweets using AI to determine their quality based on:

  1. Text Content: Evaluates writing quality, information value, and relevance
  2. Media Analysis: Uses vision-capable models to understand images in tweets
  3. Context Recognition: Considers conversations and quoted tweets for better rating
  4. Custom Criteria: Applies your personalized rating instructions

Each tweet receives a score from 1-10, which is displayed in the top-right corner. Low-scoring tweets can be automatically hidden based on your threshold setting.

Prerequisites

  • An OpenRouter API key
  • A userscript manager extension
  • Twitter/X account

Limitations

  • API call limits based on your OpenRouter plan
  • Only works on Twitter/X web interface, not the mobile app
  • Processing tweets may take a moment depending on the model selected

Contributing

Contributions are welcome! Feel free to submit issues or pull requests.

License

This userscript is provided as-is under the MIT License.

Acknowledgments

  • Thanks to OpenRouter for providing model access
  • Inspired by the need for a better social media experience