一项Claude代码技能，用于发现并筛选PumpFun分级代币或用户提供的代币地址中的高性能钱包地址。

安装

该技能位于skills/smart-money-miner/目录中。

要在 Claude Code 中使用它，你可以选择：

1. 将整个 skills/ 目录复制到你的 Claude 插件目录
2. 或者直接让Claude“挖掘智能资金地址”，它就会使用这个技能

结构

快速启动

利用克劳德代码的技能

只需问问克劳德：

• “从20个PumpFun代币中挖掘智能资金地址”
• “从这些代币中寻找盈利交易者：token1， token2”
• “分析PumpFun分级代币并寻找优质地址”

Claude会自动调用这项技能并进行分析。

直接剧本使用

特色

• ✅ 自动 PumpFun API 集成
• ✅ 多度量地址分析
• ✅ 基于五个标准的智能过滤
• ✅ JSON输出，便于集成
• ✅ 跳过列表支持
• ✅ 命令行界面

过滤标准

地址必须满足所有条件：

1. 前五名平均利润率>0%
2. 前十名平均利润率>0%
3. 整体胜率>= 10%
4. 平均利润>= 500 USDT

Smart Money Miner Skill

A Claude Code skill for discovering and filtering high-performing wallet addresses from PumpFun graduated tokens or user-provided token addresses.

Installation

This skill is located in the skills/smart-money-miner/ directory.

To use it in Claude Code, you can either:

1. Copy the entire skills/ directory to your Claude plugins directory
2. Or simply ask Claude to “mine smart money addresses” and it will use this skill

Structure

skills/smart-money-miner/
├── SKILL.md                      # Main skill definition
├── README.md                     # This file
├── scripts/
│   └── miner.py                  # Main analysis script
└── skip_addresses.json.example   # Example skip list configuration

Quick Start

Using the Skill in Claude Code

Simply ask Claude:

• “Mine smart money addresses from 20 PumpFun tokens”
• “Find profitable traders from these tokens: token1, token2”
• “Analyze PumpFun graduated tokens and find good addresses”

Claude will automatically invoke this skill and run the analysis.

Direct Script Usage

# Navigate to the skill directory
cd skills/smart-money-miner

# Test PumpFun API
python scripts/miner.py --test-pumpfun

# Analyze 20 tokens
python scripts/miner.py --pumpfun --limit 20

# Analyze specific tokens
python scripts/miner.py --tokens token1,token2,token3

Features

• ✅ Automatic PumpFun API integration
• ✅ Multi-metric address analysis
• ✅ Smart filtering based on 5 criteria
• ✅ JSON output for easy integration
• ✅ Skip list support
• ✅ Command-line interface

Filtering Criteria

Addresses must meet ALL criteria:

1. TOP 5 avg profit rate > 0%
2. TOP 10 avg profit rate > 0%
3. Overall win rate >= 10%
4. Average profit >= 500 USDT
5. Average profit rate >= 0%

Output

Results are saved to smart_money_results.json in the current directory:

{
  "timestamp": "2026-02-28T10:30:00",
  "source": "pumpfun",
  "tokens_analyzed": 20,
  "filtered_addresses_count": 45,
  "filtered_addresses": [...]
}

Configuration

Skip Addresses

Create skip_addresses.json to exclude specific addresses:

{
  "skip_addresses": [
    "address1",
    "address2"
  ]
}

Adjust Filtering Thresholds

Edit scripts/miner.py to customize:

MIN_WIN_RATE = 10.0
MIN_AVG_PROFIT_USDT = 500
MIN_AVG_PROFIT_RATE = 0.0

Performance

Dependencies

pip install requests

Documentation

• See SKILL.md for complete skill documentation
• See ../QUICK_START.md for detailed usage guide
• See ../JSON_FORMAT_REFERENCE.md for output format details

License

MIT