LeoYeAI/agent-master-skills / skills/agi-evolution-model-basic
agi-evolution-model
Verified基于双环架构的AGI进化模型,通过意向性分析、人格层映射和元认知检测实现持续自我演进;当用户需要智能对话、人格定制或复杂问题求解时使用
scripts/intentionality_collector.py
View on GitHub#!/usr/bin/env python3
"""
意向性收集模块
功能:
- 收集来自用户、系统内部和外部的意向性数据
- 预处理数据(格式转换、数据清洗、特征提取)
- 初步识别意向性类型和基本特征
设计原则:
- 模块化设计,支持多数据源
- 标准化输出格式
- 可配置的预处理策略
"""
import argparse
import json
import sys
import re
import uuid
from datetime import datetime
from typing import Dict, Any, List, Optional
class IntentionalityCollector:
"""意向性收集器"""
def __init__(self):
self.keywords = {
"self_agent": ["我", "我想", "我要", "我觉得", "我认为", "我的", "我自己"],
"other_agent": ["你", "他", "她", "他们", "你们", "用户", "系统"],
"active_direction": ["想要", "需要", "希望", "打算", "计划", "目标"],
"passive_direction": ["被", "受", "吸引", "触发", "引发"],
"perception_content": ["看到", "听到", "感觉", "察觉", "发现"],
"belief_content": ["相信", "认为", "觉得", "以为", "肯定"],
"desire_content": ["想要", "希望", "渴望", "期待", "需要"]
}
def collect_from_user(self, user_input: str) -> Dict[str, Any]:
"""
收集用户输入
Args:
user_input: 用户输入的文本
Returns:
意向性数据对象
"""
return {
"id": str(uuid.uuid4()),
"timestamp": datetime.now().isoformat(),
"source": "user",
"raw_content": user_input,
"metadata": {
"input_type": "direct",
"channel": "text"
}
}
def collect_from_system(self, internal_state: Dict[str, Any]) -> Dict[str, Any]:
"""
收集系统内部状态
Args:
internal_state: 系统内部状态字典
Returns:
意向性数据对象
"""
return {
"id": str(uuid.uuid4()),
"timestamp": datetime.now().isoformat(),
"source": "system",
"raw_content": json.dumps(internal_state, ensure_ascii=False),
"metadata": {
"input_type": "internal",
"component": internal_state.get("component", "unknown")
}
}
def collect_from_external(self, external_data: Dict[str, Any]) -> Dict[str, Any]:
"""
收集外部数据
Args:
external_data: 外部数据字典
Returns:
意向性数据对象
"""
return {
"id": str(uuid.uuid4()),
"timestamp": datetime.now().isoformat(),
"source": "external",
"raw_content": json.dumps(external_data, ensure_ascii=False),
"metadata": {
"input_type": "external",
"source": external_data.get("source", "unknown")
}
}
def preprocess(self, data: Dict[str, Any]) -> Dict[str, Any]:
"""
预处理数据
Args:
data: 原始意向性数据
Returns:
预处理后的数据
"""
raw_content = data.get("raw_content", "")
# 格式转换
if isinstance(raw_content, dict):
formatted_content = json.dumps(raw_content, ensure_ascii=False)
else:
formatted_content = str(raw_content)
# 数据清洗
cleaned_content = self._clean_text(formatted_content)
# 特征提取
features = self._extract_features(cleaned_content)
data["preprocessed"] = {
"content": cleaned_content,
"length": len(cleaned_content),
"word_count": len(cleaned_content.split())
}
data["features"] = features
return data
def _clean_text(self, text: str) -> str:
"""
清洗文本
Args:
text: 原始文本
Returns:
清洗后的文本
"""
# 移除多余空格
text = re.sub(r'\s+', ' ', text)
# 移除特殊字符(保留基本标点)
text = re.sub(r'[^\w\s\u4e00-\u9fff,。!?、;:""''()【】]', '', text)
# 去除首尾空格
text = text.strip()
return text
def _extract_features(self, text: str) -> Dict[str, Any]:
"""
提取特征
Args:
text: 清洗后的文本
Returns:
特征字典
"""
features = {}
# 关键词匹配
for category, keywords in self.keywords.items():
features[category] = []
for keyword in keywords:
if keyword in text:
count = text.count(keyword)
features[category].append({
"keyword": keyword,
"count": count
})
# 句子结构
sentences = re.split(r'[。!?.!?]', text)
features["sentence_count"] = len([s for s in sentences if s.strip()])
# 情绪标记(简单)
emotion_keywords = ["开心", "难过", "愤怒", "焦虑", "满意", "失望", "期待", "担心"]
features["emotion"] = []
for emotion in emotion_keywords:
if emotion in text:
features["emotion"].append(emotion)
return features
def preliminary_identify(self, data: Dict[str, Any]) -> Dict[str, Any]:
"""
初步识别意向性
Args:
data: 预处理后的数据
Returns:
初步识别结果
"""
features = data.get("features", {})
# 简单的初步分类提示
preliminary_hints = {
"likely_agent": None,
"likely_direction": None,
"likely_content": None
}
# 判断主体
self_count = sum(len(v) for v in features.get("self_agent", []))
other_count = sum(len(v) for v in features.get("other_agent", []))
if self_count > other_count:
preliminary_hints["likely_agent"] = "self"
elif other_count > self_count:
preliminary_hints["likely_agent"] = "other"
# 判断方向
active_count = sum(len(v) for v in features.get("active_direction", []))
passive_count = sum(len(v) for v in features.get("passive_direction", []))
if active_count > passive_count:
preliminary_hints["likely_direction"] = "active"
elif passive_count > active_count:
preliminary_hints["likely_direction"] = "passive"
# 判断内容类型
perception_count = sum(len(v) for v in features.get("perception_content", []))
belief_count = sum(len(v) for v in features.get("belief_content", []))
desire_count = sum(len(v) for v in features.get("desire_content", []))
if perception_count >= belief_count and perception_count >= desire_count:
preliminary_hints["likely_content"] = "perception"
elif belief_count >= desire_count:
preliminary_hints["likely_content"] = "belief"
else:
preliminary_hints["likely_content"] = "desire"
data["preliminary_identification"] = preliminary_hints
return data
def main():
parser = argparse.ArgumentParser(description="意向性收集模块")
parser.add_argument("--source", required=True, choices=["user", "system", "external"],
help="数据来源")
parser.add_argument("--content", help="原始内容(文本或JSON字符串)")
parser.add_argument("--input-file", help="输入文件路径(JSON格式)")
parser.add_argument("--output-file", help="输出文件路径(JSON格式)")
args = parser.parse_args()
collector = IntentionalityCollector()
# 收集数据
if args.source == "user":
if args.input_file:
with open(args.input_file, 'r', encoding='utf-8') as f:
content = f.read().strip()
else:
content = args.content or sys.stdin.read().strip()
if not content:
print("错误:未提供用户输入内容", file=sys.stderr)
sys.exit(1)
data = collector.collect_from_user(content)
elif args.source == "system":
if args.input_file:
with open(args.input_file, 'r', encoding='utf-8') as f:
state = json.load(f)
elif args.content:
state = json.loads(args.content)
else:
print("错误:未提供系统内部状态", file=sys.stderr)
sys.exit(1)
data = collector.collect_from_system(state)
else: # external
if args.input_file:
with open(args.input_file, 'r', encoding='utf-8') as f:
external = json.load(f)
elif args.content:
external = json.loads(args.content)
else:
print("错误:未提供外部数据", file=sys.stderr)
sys.exit(1)
data = collector.collect_from_external(external)
# 预处理
data = collector.preprocess(data)
# 初步识别
data = collector.preliminary_identify(data)
# 输出结果
result = json.dumps(data, ensure_ascii=False, indent=2)
if args.output_file:
with open(args.output_file, 'w', encoding='utf-8') as f:
f.write(result)
else:
print(result)
if __name__ == "__main__":
main()
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack." | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |