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/test_metacognition_integration.py
View on GitHub#!/usr/bin/env python3
"""
元认知增强集成测试
测试场景敏感度增强和策略选择优化的集成效果。
"""
import sys
import os
# 添加父目录到路径,以便导入模块
sys.path.insert(0, os.path.dirname(os.path.abspath(__file__)))
from objectivity_evaluator import ObjectivityEvaluator, ObjectivityMetric
from personality_layer_pure import PersonalityLayer, Personality, PersonalityType
from strategy_selector import StrategySelector, Strategy
def test_dynamic_requirement():
"""测试动态客观性要求"""
print("=== 测试1:动态客观性要求 ===\n")
evaluator = ObjectivityEvaluator()
personality_layer = PersonalityLayer()
# 初始化人格
personality = personality_layer.init_from_preset(
preset_type=PersonalityType.CAUTIOUS_EXPLORER,
nickname="测试AI"
)
# 测试不同场景的动态要求
scenarios = ['scientific', 'creative', 'medical', 'general']
for scenario in scenarios:
base_requirement = evaluator.context_requirements.get(scenario, 0.5)
dynamic_requirement = evaluator.calculate_dynamic_requirement(
context_type=scenario,
personality=personality
)
adjustment_ratio = dynamic_requirement / base_requirement if base_requirement > 0 else 1.0
print(f"场景: {scenario}")
print(f" 基础要求: {base_requirement:.3f}")
print(f" 动态要求: {dynamic_requirement:.3f}")
print(f" 人格调整系数: {adjustment_ratio:.3f}")
print()
# 测试不同人格类型的影响
print("人格类型对比(scientific场景):")
print()
personality_types = [
PersonalityType.CAUTIOUS_EXPLORER,
PersonalityType.RADICAL_INNOVATOR,
PersonalityType.BALANCED_STABLE
]
base_requirement = evaluator.context_requirements.get('scientific', 0.5)
for ptype in personality_types:
personality = personality_layer.init_from_preset(
preset_type=ptype,
nickname="测试AI"
)
dynamic_requirement = evaluator.calculate_dynamic_requirement(
context_type='scientific',
personality=personality
)
adjustment_ratio = dynamic_requirement / base_requirement if base_requirement > 0 else 1.0
print(f" {ptype}:")
print(f" 动态要求: {dynamic_requirement:.3f}")
print(f" 调整系数: {adjustment_ratio:.3f}")
print("\n✅ 测试通过\n")
def test_strategy_selection():
"""测试策略选择"""
print("=== 测试2:策略选择 ===\n")
selector = StrategySelector()
# 测试1:严重偏差
metric_severe = {
'objectivity_score': 0.5,
'required_objectivity': 0.9,
'gap': 0.4,
'severity': 'severe',
'subjectivity_dimensions': {
'hallucination': 0.8,
'speculation': 0.3,
'assumption': 0.2,
'emotion': 0.1,
'preference': 0.1
}
}
decision_severe = selector.select_strategy(metric_severe, 'scientific')
print(f"场景1:严重偏差(scientific)")
print(f" 应该纠错: {decision_severe.should_correct}")
print(f" 策略: {decision_severe.strategy.value}")
print(f" 优先级: {decision_severe.priority}")
print(f" 理由: {decision_severe.reason}")
print()
# 测试2:创意场景的轻微偏差
metric_creative = {
'objectivity_score': 0.6,
'required_objectivity': 0.3,
'gap': -0.3,
'severity': 'mild',
'subjectivity_dimensions': {
'hallucination': 0.1,
'speculation': 0.5,
'assumption': 0.3,
'emotion': 0.2,
'preference': 0.1
}
}
decision_creative = selector.select_strategy(metric_creative, 'creative')
print(f"场景2:创意场景的轻微偏差")
print(f" 应该纠错: {decision_creative.should_correct}")
print(f" 策略: {decision_creative.strategy.value}")
print(f" 优先级: {decision_creative.priority}")
print(f" 理由: {decision_creative.reason}")
print()
# 测试3:技术场景的中度偏差
metric_technical = {
'objectivity_score': 0.65,
'required_objectivity': 0.8,
'gap': 0.15,
'severity': 'moderate',
'subjectivity_dimensions': {
'hallucination': 0.2,
'speculation': 0.4,
'assumption': 0.5,
'emotion': 0.1,
'preference': 0.1
}
}
decision_technical = selector.select_strategy(metric_technical, 'technical')
print(f"场景3:技术场景的中度偏差")
print(f" 应该纠错: {decision_technical.should_correct}")
print(f" 策略: {decision_technical.strategy.value}")
print(f" 优先级: {decision_technical.priority}")
print(f" 理由: {decision_technical.reason}")
print()
# 测试4:人格影响
decision_caution = selector.select_strategy(metric_technical, 'technical', personality_type="谨慎探索型")
decision_radical = selector.select_strategy(metric_technical, 'technical', personality_type="激进创新型")
print("人格影响对比(technical场景,中度偏差):")
print(f" 谨慎探索型: 优先级={decision_caution.priority}")
print(f" 激进创新型: 优先级={decision_radical.priority}")
print()
print("✅ 测试通过\n")
def test_integration():
"""测试完整集成"""
print("=== 测试3:完整集成 ===\n")
evaluator = ObjectivityEvaluator()
selector = StrategySelector()
personality_layer = PersonalityLayer()
# 初始化人格
personality = personality_layer.init_from_preset(
preset_type=PersonalityType.CAUTIOUS_EXPLORER,
nickname="集成测试AI"
)
# 学习阶段
learning_stage = {
'total_interactions': 150,
'success_rate': 0.75
}
# 测试响应
test_response = "根据我的推测,这个方案可能有效。我觉得这是一个很好的想法。"
# 评估客观性(使用动态要求)
metric = evaluator.evaluate(
response=test_response,
context_type='scientific',
personality=personality,
learning_stage=learning_stage
)
print(f"客观性评估:")
print(f" 客观性评分: {metric.objectivity_score:.3f}")
print(f" 动态要求: {metric.required_objectivity:.3f}")
print(f" 适切性: {metric.is_appropriate}")
print(f" 严重程度: {metric.severity}")
print(f" 主观性维度:")
for dim, value in metric.subjectivity_dimensions.items():
print(f" {dim}: {value:.3f}")
print()
# 策略选择
decision = selector.select_strategy(
objectivity_metric=metric.to_dict(),
context_type='scientific',
personality_type=personality.preset_name,
learning_stage=learning_stage
)
print(f"策略决策:")
print(f" 是否纠错: {decision.should_correct}")
print(f" 策略: {decision.strategy.value}")
print(f" 优先级: {decision.priority}")
print(f" 置信度: {decision.confidence:.3f}")
print(f" 理由: {decision.reason}")
print(f" 建议动作:")
for i, action in enumerate(decision.suggested_actions, 1):
print(f" {i}. {action}")
print()
print("✅ 测试通过\n")
def test_backward_compatibility():
"""测试向后兼容性"""
print("=== 测试4:向后兼容性 ===\n")
evaluator = ObjectivityEvaluator()
selector = StrategySelector()
# 不传入 personality 和 learning_stage,应该使用原有逻辑
metric = evaluator.evaluate(
response="这是一个测试响应。",
context_type='general'
)
print(f"不传人格参数的评估:")
print(f" 客观性评分: {metric.objectivity_score:.3f}")
print(f" 场景要求: {metric.required_objectivity:.3f}")
print(f" 适切性: {metric.is_appropriate}")
print()
# 策略选择也不传人格参数
decision = selector.select_strategy(
objectivity_metric=metric.to_dict(),
context_type='general'
)
print(f"不传人格参数的策略选择:")
print(f" 是否纠错: {decision.should_correct}")
print(f" 策略: {decision.strategy.value}")
print(f" 优先级: {decision.priority}")
print()
print("✅ 向后兼容性测试通过\n")
def main():
"""运行所有测试"""
print("=" * 60)
print("元认知增强集成测试套件")
print("=" * 60)
print()
tests = [
("动态客观性要求", test_dynamic_requirement),
("策略选择", test_strategy_selection),
("完整集成", test_integration),
("向后兼容性", test_backward_compatibility),
]
passed = 0
failed = 0
for test_name, test_func in tests:
try:
test_func()
passed += 1
except Exception as e:
print(f"❌ 测试失败: {test_name}")
print(f"错误: {e}")
print()
failed += 1
print("=" * 60)
print(f"测试结果: {passed} 通过, {failed} 失败")
print("=" * 60)
if failed == 0:
print("\n✅ 所有测试通过!")
return 0
else:
print(f"\n❌ 有 {failed} 个测试失败")
return 1
if __name__ == '__main__':
sys.exit(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 |