LeoYeAI/agent-master-skills / skills/anima-aios
anima-aios
VerifiedAn AI Agent cognitive growth system built on the native OpenClaw architecture. It provides agents with persistent memory management, visual intimacy progression, a 5-dimensional cognitive profile, gamified daily quests, team leaderboards, and a 5-layer memory architecture with Knowledge Palace, Pyramid thinking, and Ebbinghaus decay function. 基于 OpenClaw 原生架构的 AI Agent 认知成长体系,为 Agent 提供五层记忆架构、知识宫殿、金字塔知识组织、记忆衰减函数、LLM 智能处理、永久化记忆管理、可视化亲密度成长、五维认知画像、游戏化每日任务和团队排行榜。
health/__init__.py
View on GitHub#!/usr/bin/env python3
# -*- coding: utf-8 -*-
# Copyright 2026 Anima-AIOS Contributors
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
"""
Anima AIOS v6.0 - Health System
健康系统,包含 5 个子模块:
- manager: 总调度(Doctor 入口)
- hygiene: 数据卫生(完整性 + 去重 + 清理)
- correction: 自动纠错
- evolution: 每日进化(凌晨自动提炼)
- abstraction: 知识抽象
移植自 Z 的 health/ 目录(6 个 JS 文件 → 1 个 Python 模块)。
Author: 清禾
Date: 2026-03-24
Version: 6.0.0
"""
import os
import sys
import json
import logging
from pathlib import Path
from datetime import datetime
from typing import Dict, List, Optional
# core 模块通过相对导入引用(在各方法内按需导入)
logger = logging.getLogger(__name__)
class HealthHygiene:
"""数据卫生检查:完整性 + 去重 + 清理"""
def __init__(self, agent_name: str, facts_base: str = None):
self.agent_name = agent_name
self.base = Path(facts_base) / agent_name
def check_integrity(self) -> Dict:
"""检查数据完整性"""
issues = []
# 检查核心目录
required_dirs = [
"facts/episodic", "facts/semantic",
"palace", "palace/rooms", "pyramid",
"decay", "health"
]
for d in required_dirs:
path = self.base / d
if not path.exists():
issues.append({"type": "missing_dir", "path": str(path), "severity": "high"})
# 检查核心文件
required_files = ["exp_history.jsonl"]
for f in required_files:
path = self.base / f
if not path.exists():
issues.append({"type": "missing_file", "path": str(path), "severity": "medium"})
# 检查 facts 格式
for fact_type in ["episodic", "semantic"]:
fact_dir = self.base / "facts" / fact_type
if not fact_dir.exists():
continue
for md_file in fact_dir.glob("*.md"):
content = md_file.read_text(encoding='utf-8')
if not content.startswith("---"):
issues.append({
"type": "malformed_fact",
"path": str(md_file),
"severity": "low",
"detail": "缺少 frontmatter"
})
return {
"status": "healthy" if not issues else "issues_found",
"issues": issues,
"checked_at": datetime.now().isoformat()
}
def find_duplicates(self) -> List[Dict]:
"""查找重复的 facts"""
seen_hashes = {}
duplicates = []
for fact_type in ["episodic", "semantic"]:
fact_dir = self.base / "facts" / fact_type
if not fact_dir.exists():
continue
for md_file in fact_dir.glob("*.md"):
content = md_file.read_text(encoding='utf-8')
# 取正文部分的哈希
parts = content.split("---", 2)
body = parts[2].strip() if len(parts) >= 3 else content.strip()
import hashlib
h = hashlib.md5(body.encode()).hexdigest()
if h in seen_hashes:
duplicates.append({
"file": str(md_file),
"duplicate_of": seen_hashes[h],
"hash": h
})
else:
seen_hashes[h] = str(md_file)
return duplicates
def cleanup_empty(self) -> int:
"""清理空文件"""
removed = 0
for fact_type in ["episodic", "semantic"]:
fact_dir = self.base / "facts" / fact_type
if not fact_dir.exists():
continue
for md_file in fact_dir.glob("*.md"):
if md_file.stat().st_size == 0:
md_file.unlink()
removed += 1
return removed
class HealthCorrection:
"""自动纠错:检测并修复常见数据问题"""
def __init__(self, agent_name: str, facts_base: str = None):
self.agent_name = agent_name
self.base = Path(facts_base) / agent_name
def fix_missing_dirs(self) -> int:
"""创建缺失的目录"""
fixed = 0
required = [
"facts/episodic", "facts/semantic",
"palace", "palace/rooms", "pyramid",
"decay", "health"
]
for d in required:
path = self.base / d
if not path.exists():
path.mkdir(parents=True, exist_ok=True)
fixed += 1
return fixed
def fix_exp_history(self) -> bool:
"""修复损坏的 exp_history.jsonl"""
exp_file = self.base / "exp_history.jsonl"
if not exp_file.exists():
exp_file.touch()
return True
# 验证每行是合法 JSON
lines = exp_file.read_text(encoding='utf-8').splitlines()
valid_lines = []
fixed = False
for line in lines:
line = line.strip()
if not line:
continue
try:
json.loads(line)
valid_lines.append(line)
except json.JSONDecodeError:
fixed = True # 跳过损坏行
if fixed:
exp_file.write_text('\n'.join(valid_lines) + '\n', encoding='utf-8')
return fixed
def run_all_fixes(self) -> Dict:
"""执行所有自动修复"""
return {
"dirs_fixed": self.fix_missing_dirs(),
"exp_fixed": self.fix_exp_history(),
"timestamp": datetime.now().isoformat()
}
class HealthEvolution:
"""每日进化:自动提炼高价值记忆"""
def __init__(self, agent_name: str, facts_base: str = None):
self.agent_name = agent_name
if facts_base is None:
try:
from ..config.path_config import get_config
except ImportError:
from config.path_config import get_config
facts_base = str(get_config().facts_base)
self.facts_base = facts_base
self.log_file = Path(facts_base) / agent_name / "health" / "evolution-log.jsonl"
self.log_file.parent.mkdir(parents=True, exist_ok=True)
def run_daily(self) -> Dict:
"""
执行每日进化(设计为凌晨 3:00 由 cron 触发)
流程:
1. 运行提炼引擎(L2→L3)
2. 运行宫殿分类
3. 运行金字塔提炼
4. 记录日志
"""
stats = {"timestamp": datetime.now().isoformat(), "steps": {}}
# Step 1: L2→L3 提炼
try:
try:
from ..core.distill_engine import DistillEngine
except ImportError:
from distill_engine import DistillEngine
engine = DistillEngine(self.agent_name, self.facts_base)
distill_result = engine.run(batch_size=50)
stats["steps"]["distill"] = distill_result
except Exception as e:
stats["steps"]["distill"] = {"error": str(e)}
# Step 2: 宫殿分类
try:
try:
from ..core.palace_classifier import PalaceClassifier
except ImportError:
from palace_classifier import PalaceClassifier
classifier = PalaceClassifier(self.agent_name, self.facts_base)
unclassified = classifier.get_unclassified()
if unclassified:
classify_result = classifier.classify_batch(unclassified)
stats["steps"]["classify"] = classify_result
else:
stats["steps"]["classify"] = {"message": "无待分类知识"}
except Exception as e:
stats["steps"]["classify"] = {"error": str(e)}
# Step 3: 金字塔自动提炼
try:
try:
from ..core.pyramid_engine import PyramidEngine
except ImportError:
from pyramid_engine import PyramidEngine
pyramid = PyramidEngine(self.agent_name, self.facts_base, auto_distill=True)
topics = pyramid.get_topics()
distilled_topics = 0
for topic, level_counts in topics.items():
if level_counts.get("instances", 0) >= 3 and level_counts.get("rules", 0) == 0:
pyramid.distill_up(topic, "instances")
distilled_topics += 1
stats["steps"]["pyramid"] = {"topics_distilled": distilled_topics}
except Exception as e:
stats["steps"]["pyramid"] = {"error": str(e)}
# 记录日志
with open(self.log_file, 'a', encoding='utf-8') as f:
f.write(json.dumps(stats, ensure_ascii=False) + '\n')
return stats
class HealthAbstraction:
"""知识抽象:举一反三,从已有知识中发现新关联"""
def __init__(self, agent_name: str, facts_base: str = None,
llm_config: Dict = None):
self.agent_name = agent_name
if facts_base is None:
try:
from ..config.path_config import get_config
except ImportError:
from config.path_config import get_config
facts_base = str(get_config().facts_base)
self.facts_base = facts_base
self.llm = None
if llm_config:
try:
try:
from ..core.distill_engine import LLMClient
except ImportError:
from distill_engine import LLMClient
self.llm = LLMClient(llm_config)
except ImportError:
pass
def find_connections(self, limit: int = 5) -> List[Dict]:
"""
发现跨房间的知识关联
"""
try:
try:
from ..core.palace_index import PalaceIndex
except ImportError:
from palace_index import PalaceIndex
palace = PalaceIndex(self.agent_name, self.facts_base)
except ImportError:
return []
rooms = palace.list_rooms()
all_items = []
for room in rooms:
if room["id"] == "_inbox":
continue
items = palace.get_room_items(room["id"])
for item in items:
item["room_id"] = room["id"]
item["room_name"] = room["name"]
all_items.append(item)
if len(all_items) < 2:
return []
connections = []
# 简单关联:不同房间中有相同标签的知识
from collections import defaultdict
tag_map = defaultdict(list)
for item in all_items:
for tag in item.get("tags", []):
tag_map[tag].append(item)
for tag, items in tag_map.items():
rooms_involved = set(i["room_id"] for i in items)
if len(rooms_involved) >= 2:
connections.append({
"type": "shared_tag",
"tag": tag,
"rooms": list(rooms_involved),
"items": [i["fact_id"] for i in items]
})
return connections[:limit]
class HealthManager:
"""
健康系统总调度器
整合 5 个子模块,提供统一的诊断和修复接口。
Doctor 命令的后端。
"""
def __init__(self, agent_name: str, facts_base: str = None,
llm_config: Dict = None):
self.agent_name = agent_name
if facts_base is None:
try:
from ..config.path_config import get_config
except ImportError:
from config.path_config import get_config
facts_base = str(get_config().facts_base)
self.facts_base = facts_base
self.hygiene = HealthHygiene(agent_name, facts_base)
self.correction = HealthCorrection(agent_name, facts_base)
self.evolution = HealthEvolution(agent_name, facts_base)
self.abstraction = HealthAbstraction(agent_name, facts_base, llm_config)
def full_checkup(self) -> Dict:
"""完整健康检查"""
report = {
"agent": self.agent_name,
"timestamp": datetime.now().isoformat(),
"modules": {}
}
# 1. 完整性检查
report["modules"]["integrity"] = self.hygiene.check_integrity()
# 2. 重复检查
dups = self.hygiene.find_duplicates()
report["modules"]["duplicates"] = {
"count": len(dups),
"items": dups[:10]
}
# 3. 衰减状态
try:
try:
from ..core.decay_function import DecayManager
except ImportError:
from decay_function import DecayManager
try:
from ..core.fact_store import FactStore
except ImportError:
from fact_store import FactStore
store = FactStore(self.agent_name, self.facts_base)
decay = DecayManager(self.agent_name, self.facts_base)
all_facts = []
for ft in ["episodic", "semantic"]:
for f in store.list_facts(ft, limit=100):
all_facts.append({"fact_id": f.fact_id, "quality": f.quality, "content": f.content})
decay_stats = decay.scan_all(all_facts)
report["modules"]["decay"] = {
"total": decay_stats["total"],
"healthy": len(decay_stats["healthy"]),
"weakening": len(decay_stats["weakening"]),
"needs_review": len(decay_stats["needs_review"]),
"forgetting": len(decay_stats["forgetting"]),
"archivable": len(decay_stats["archivable"])
}
except Exception as e:
report["modules"]["decay"] = {"error": str(e)}
# 4. 宫殿状态
try:
try:
from ..core.palace_index import PalaceIndex
except ImportError:
from palace_index import PalaceIndex
palace = PalaceIndex(self.agent_name, self.facts_base)
report["modules"]["palace"] = palace.get_stats()
except Exception as e:
report["modules"]["palace"] = {"error": str(e)}
# 5. 金字塔状态
try:
try:
from ..core.pyramid_engine import PyramidEngine
except ImportError:
from pyramid_engine import PyramidEngine
pyramid = PyramidEngine(self.agent_name, self.facts_base)
report["modules"]["pyramid"] = pyramid.get_stats()
except Exception as e:
report["modules"]["pyramid"] = {"error": str(e)}
# 6. 知识关联
connections = self.abstraction.find_connections()
report["modules"]["connections"] = {
"count": len(connections),
"items": connections
}
# 总���
issues = report["modules"]["integrity"].get("issues", [])
high_issues = [i for i in issues if i.get("severity") == "high"]
report["overall"] = "healthy" if not high_issues else "needs_attention"
return report
def auto_fix(self) -> Dict:
"""执行自动修复"""
return {
"corrections": self.correction.run_all_fixes(),
"empty_cleaned": self.hygiene.cleanup_empty(),
"timestamp": datetime.now().isoformat()
}
def run_evolution(self) -> Dict:
"""触发每日进化"""
return self.evolution.run_daily()
def main():
import argparse
logging.basicConfig(
level=logging.INFO,
format='%(asctime)s [Health] %(levelname)s: %(message)s'
)
parser = argparse.ArgumentParser(description='Anima Health System')
parser.add_argument('--agent', type=str, default='')
parser.add_argument('--facts-base', type=str, default='/home/画像')
parser.add_argument('action', choices=['checkup', 'fix', 'evolve'], help='操作')
args = parser.parse_args()
agent = args.agent or os.getenv("ANIMA_AGENT_NAME", "unknown")
mgr = HealthManager(agent, args.facts_base)
if args.action == 'checkup':
report = mgr.full_checkup()
print(json.dumps(report, ensure_ascii=False, indent=2))
elif args.action == 'fix':
result = mgr.auto_fix()
print(json.dumps(result, ensure_ascii=False, indent=2))
elif args.action == 'evolve':
result = mgr.run_evolution()
print(json.dumps(result, ensure_ascii=False, indent=2))
if __name__ == '__main__':
main()
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 323k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 323k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 323k | Updated | Claude CodeCodex | Verified | Install |