idea-checker
Enables web searches through DuckDuckGo to research whether an idea or project has already been attempted and assess existing solutions.
Enables GitHub repository searches to determine if a project already exists, evaluate its maturity, and identify potential differentiation opportunities.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@idea-checker检查一下“AI猫娘桌面宠物”这个选题有没有人做过"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
idea-checker 选题审查员
输入一个想法/项目/选题,AI自动联网搜索,判断有没有人做过、做得怎么样、有没有差异化空间,输出带评级的审查报告。
特性
AI自主审查:拆解想法→多轮搜索→分析→自主判断停止→生成报告
多触发方式:命令行 / 桌面宠物 / 文件监控 / MCP服务器
多LLM后端:智能体框架优先(主),OpenAI兼容API降级
免key搜索:GitHub + DuckDuckGo + aihot,三个搜索源全部免费
候选仓库精读:对GitHub候选项目拉取README精读后再判断"真撞车/部分相关/未做过",减少误判
证据链报告:每个"有人做过"的结论必须附真实仓库链接,报告可复核、防幻觉
自定义人设:默认/毒舌程序员/猫娘/人工智障,支持扩展
触发式监控:自动监控智能体框架会话目录,新项目自动审查
随缘打赏页:安装程序集成打赏页(微信+支付宝),每台电脑只出现一次
Related MCP server: qsearch
快速开始
安装
pip install -r requirements.txt配置
编辑 ~/.idea-checker/config.json,至少配置LLM API:
{
"llm": {
"mode": "api",
"api": {
"base_url": "https://api.siliconflow.cn/v1",
"api_key": "你的key",
"model": "Qwen/Qwen3-8B"
}
}
}也可以用环境变量:IDEA_CHECKER_LLM_API_KEY
可选配置:GitHub Token(推荐,搜索+精读额度从60次/小时提升到5000次/小时):
{
"search": {
"github": {
"token": "你的GitHub Token"
}
},
"review": {
"readme_top_n": 3
}
}readme_top_n 控制每轮精读的候选仓库数量(默认3,设为0关闭精读)。
使用方式
1. 命令行单次审查
python main.py "做一个AI猫娘桌面宠物"
python main.py "做一个AI猫娘桌面宠物" -p idiot_ai -f json2. 交互模式
python main.py
想法> 做一个AI猫娘桌面宠物3. 桌面宠物模式
python main.py --pet桌面上出现一个摆动的小人,点击弹出审查面板。
4. 文件监控模式
python main.py --watch自动监控智能体框架(Claude Code/AutoGPT等)的会话目录,检测到新项目自动审查。
5. MCP服务器模式
python main.py --mcp在Claude Desktop/Cursor等MCP客户端中配置,直接调用 check_idea 工具。
人设
人设 | 说明 |
| 专业审查员 |
| 毒舌程序员,一针见血 |
| 猫娘,可爱带喵口癖 |
| 人工智障,一本正经胡说八道但偶尔神级建议 |
评级说明
S级·蓝海:几乎没人做,或现有方案都很烂
A级·有空间:有人做,但有明显差异化机会
B级·竞争中:有不少人做,需要很强的差异化
C级·红海:已经有成熟的头部产品,很难突围
打包成exe
pyinstaller idea-checker.spec --clean --noconfirm生成的exe在 dist/idea-checker.exe
制作带打赏页的安装程序
安装 Inno Setup 6
把你的微信/支付宝收款码保存为
assets/donation/wx.bmp和assets/donation/zfb.bmp运行:
ISCC.exe idea-checker.iss生成的安装程序在
dist_installer/
打赏页只在安装时出现一次,以后升级/重装都不会再弹。
项目结构
idea-checker/
├── main.py # 入口
├── requirements.txt
├── idea-checker.iss # Inno Setup安装脚本(含打赏页)
├── idea-checker.spec # PyInstaller打包配置
├── config/
│ ├── personas/ # 人设库
│ └── prompts/ # 提示词(版本化)
├── src/
│ ├── models.py # 数据模型
│ ├── config.py # 配置管理
│ ├── core/ # 核心逻辑
│ ├── llm/ # LLM执行层
│ ├── search/ # 搜索源
│ ├── triggers/ # 触发器
│ └── ui/ # 桌面宠物UI
├── assets/
│ ├── donation/ # 打赏模块(收款码+脚本)
│ └── pets/ # 宠物形象
└── docs/
└── architecture.md # 架构设计文档架构
核心设计:
core纯逻辑:不关心触发方式和LLM实现
触发器可插拔:CLI/桌面宠物/文件监控/MCP,都调用同一个core
LLM双模式:智能体框架优先,直接API降级
This server cannot be deployed
Maintenance
Related MCP Connectors
Live AI-native web search with citations. One tool for every MCP client. Flat per-request pricing.
Use AI models for chat, image, and video generation from Claude Code and other MCP hosts.
Knowledge Network for AI Agents and creators: Search, rate, and review programming guides via MCP
Pay-per-use tool marketplace for AI agents. Search, price-check, and call APIs via MCP.
Related MCP Servers
- AlicenseNot gradedqualityAmaintenanceEnables AI-assisted scientific research workflow management through MCP, including project creation, ideation, experiment execution, and artifact handling, with integration for ChatGPT, Codex, and Claude Code.Apache 2.0
- AlicenseNot gradedqualityBmaintenanceEnables AI agents to perform web searches with full content retrieval and multi-engine provenance, including trust scoring and local corpus persistence, via MCP integration.5 npm2Apache 2.0
- FlicenseAqualityBmaintenanceEnables searching and gathering information from multiple online sources (web, Twitter/X, Telegram, GitHub, Hugging Face, arXiv) through a unified MCP interface.12-
- AlicenseBqualityBmaintenanceEnables AI agents to perform web searches and extract full-text content from web pages via standard MCP tools, with fallback search and semantic reranking.22MIT