File size: 59,757 Bytes
23804b3 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 |
"""
Advanced Reasoning Engine with Persistent Memory
Implements long-term thinking, strategic planning, and persistent memory systems
Author: Cyber-LLM Development Team
Date: August 6, 2025
Version: 2.0.0
"""
import asyncio
import json
import logging
import sqlite3
import pickle
import hashlib
from datetime import datetime, timedelta
from typing import Dict, List, Any, Optional, Tuple, Set, Union
from dataclasses import dataclass, field
from enum import Enum
import threading
import time
from collections import defaultdict, deque
from pathlib import Path
import numpy as np
# Advanced reasoning imports
from abc import ABC, abstractmethod
import uuid
import networkx as nx
import yaml
class ReasoningType(Enum):
"""Types of reasoning supported by the system"""
DEDUCTIVE = "deductive" # General to specific
INDUCTIVE = "inductive" # Specific to general
ABDUCTIVE = "abductive" # Best explanation
ANALOGICAL = "analogical" # Similarity-based
CAUSAL = "causal" # Cause-effect relationships
STRATEGIC = "strategic" # Long-term planning
COUNTERFACTUAL = "counterfactual" # What-if scenarios
META_COGNITIVE = "meta_cognitive" # Reasoning about reasoning
class MemoryType(Enum):
"""Types of memory in the system"""
WORKING = "working" # Short-term active memory
EPISODIC = "episodic" # Specific experiences
SEMANTIC = "semantic" # General knowledge
PROCEDURAL = "procedural" # Skills and procedures
STRATEGIC = "strategic" # Long-term plans and goals
@dataclass
class ReasoningStep:
"""Individual step in a reasoning chain"""
step_id: str = field(default_factory=lambda: str(uuid.uuid4()))
reasoning_type: ReasoningType = ReasoningType.DEDUCTIVE
premise: str = ""
inference_rule: str = ""
conclusion: str = ""
confidence: float = 0.0
evidence: List[str] = field(default_factory=list)
timestamp: datetime = field(default_factory=datetime.now)
dependencies: List[str] = field(default_factory=list)
@dataclass
class ReasoningChain:
"""Complete reasoning chain with multiple steps"""
chain_id: str = field(default_factory=lambda: str(uuid.uuid4()))
topic: str = ""
goal: str = ""
steps: List[ReasoningStep] = field(default_factory=list)
conclusion: str = ""
confidence: float = 0.0
start_time: datetime = field(default_factory=datetime.now)
end_time: Optional[datetime] = None
success: bool = False
metadata: Dict[str, Any] = field(default_factory=dict)
@dataclass
class MemoryEntry:
"""Entry in the persistent memory system"""
memory_id: str = field(default_factory=lambda: str(uuid.uuid4()))
memory_type: MemoryType = MemoryType.EPISODIC
content: Dict[str, Any] = field(default_factory=dict)
timestamp: datetime = field(default_factory=datetime.now)
importance: float = 0.0
access_count: int = 0
last_accessed: datetime = field(default_factory=datetime.now)
decay_rate: float = 0.1
tags: Set[str] = field(default_factory=set)
@dataclass
class StrategicPlan:
"""Long-term strategic plan with goals and milestones"""
plan_id: str = field(default_factory=lambda: str(uuid.uuid4()))
title: str = ""
description: str = ""
primary_goal: str = ""
sub_goals: List[str] = field(default_factory=list)
timeline: Dict[str, datetime] = field(default_factory=dict)
milestones: List[Dict[str, Any]] = field(default_factory=list)
success_criteria: List[str] = field(default_factory=list)
risk_factors: List[str] = field(default_factory=list)
resources_required: List[str] = field(default_factory=list)
current_status: str = "planning"
progress_percentage: float = 0.0
created_at: datetime = field(default_factory=datetime.now)
updated_at: datetime = field(default_factory=datetime.now)
class PersistentMemoryManager:
"""Advanced persistent memory system for agents"""
def __init__(self, db_path: str = "data/agent_memory.db"):
self.db_path = Path(db_path)
self.db_path.parent.mkdir(parents=True, exist_ok=True)
self.logger = logging.getLogger("persistent_memory")
# Memory organization
self.working_memory = deque(maxlen=100) # Active memories
self.memory_graph = nx.DiGraph() # Semantic relationships
self.memory_cache = {} # LRU cache for fast access
# Initialize database
self._init_database()
# Background processes
self.consolidation_thread = None
self.decay_thread = None
self._start_background_processes()
def _init_database(self):
"""Initialize the SQLite database for persistent storage"""
self.conn = sqlite3.connect(self.db_path, check_same_thread=False)
self.conn.execute("PRAGMA foreign_keys = ON")
# Memory entries table
self.conn.execute("""
CREATE TABLE IF NOT EXISTS memory_entries (
memory_id TEXT PRIMARY KEY,
memory_type TEXT NOT NULL,
content BLOB NOT NULL,
timestamp REAL NOT NULL,
importance REAL NOT NULL,
access_count INTEGER DEFAULT 0,
last_accessed REAL NOT NULL,
decay_rate REAL NOT NULL,
tags TEXT DEFAULT ''
)
""")
# Reasoning chains table
self.conn.execute("""
CREATE TABLE IF NOT EXISTS reasoning_chains (
chain_id TEXT PRIMARY KEY,
topic TEXT NOT NULL,
goal TEXT NOT NULL,
steps BLOB NOT NULL,
conclusion TEXT NOT NULL,
confidence REAL NOT NULL,
start_time REAL NOT NULL,
end_time REAL,
success BOOLEAN NOT NULL,
metadata BLOB DEFAULT ''
)
""")
# Strategic plans table
self.conn.execute("""
CREATE TABLE IF NOT EXISTS strategic_plans (
plan_id TEXT PRIMARY KEY,
title TEXT NOT NULL,
description TEXT NOT NULL,
primary_goal TEXT NOT NULL,
sub_goals BLOB NOT NULL,
timeline BLOB NOT NULL,
milestones BLOB NOT NULL,
success_criteria BLOB NOT NULL,
risk_factors BLOB NOT NULL,
resources_required BLOB NOT NULL,
current_status TEXT NOT NULL,
progress_percentage REAL NOT NULL,
created_at REAL NOT NULL,
updated_at REAL NOT NULL
)
""")
# Memory relationships table
self.conn.execute("""
CREATE TABLE IF NOT EXISTS memory_relationships (
id INTEGER PRIMARY KEY AUTOINCREMENT,
source_memory_id TEXT NOT NULL,
target_memory_id TEXT NOT NULL,
relationship_type TEXT NOT NULL,
strength REAL NOT NULL,
created_at REAL NOT NULL,
FOREIGN KEY (source_memory_id) REFERENCES memory_entries (memory_id),
FOREIGN KEY (target_memory_id) REFERENCES memory_entries (memory_id)
)
""")
self.conn.commit()
self.logger.info("Persistent memory database initialized")
async def store_memory(self, memory_entry: MemoryEntry) -> str:
"""Store a memory entry in persistent storage"""
try:
# Store in database
self.conn.execute("""
INSERT OR REPLACE INTO memory_entries
(memory_id, memory_type, content, timestamp, importance,
access_count, last_accessed, decay_rate, tags)
VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?)
""", (
memory_entry.memory_id,
memory_entry.memory_type.value,
pickle.dumps(memory_entry.content),
memory_entry.timestamp.timestamp(),
memory_entry.importance,
memory_entry.access_count,
memory_entry.last_accessed.timestamp(),
memory_entry.decay_rate,
json.dumps(list(memory_entry.tags))
))
self.conn.commit()
# Add to working memory if important
if memory_entry.importance > 0.5:
self.working_memory.append(memory_entry)
# Update cache
self.memory_cache[memory_entry.memory_id] = memory_entry
self.logger.debug(f"Stored memory: {memory_entry.memory_id}")
return memory_entry.memory_id
except Exception as e:
self.logger.error(f"Error storing memory: {e}")
return None
async def retrieve_memory(self, memory_id: str) -> Optional[MemoryEntry]:
"""Retrieve a specific memory by ID"""
# Check cache first
if memory_id in self.memory_cache:
memory = self.memory_cache[memory_id]
memory.access_count += 1
memory.last_accessed = datetime.now()
return memory
try:
cursor = self.conn.execute("""
SELECT * FROM memory_entries WHERE memory_id = ?
""", (memory_id,))
row = cursor.fetchone()
if row:
memory = MemoryEntry(
memory_id=row[0],
memory_type=MemoryType(row[1]),
content=pickle.loads(row[2]),
timestamp=datetime.fromtimestamp(row[3]),
importance=row[4],
access_count=row[5] + 1,
last_accessed=datetime.now(),
decay_rate=row[7],
tags=set(json.loads(row[8]))
)
# Update access count
self.conn.execute("""
UPDATE memory_entries
SET access_count = ?, last_accessed = ?
WHERE memory_id = ?
""", (memory.access_count, memory.last_accessed.timestamp(), memory_id))
self.conn.commit()
# Cache the memory
self.memory_cache[memory_id] = memory
return memory
except Exception as e:
self.logger.error(f"Error retrieving memory {memory_id}: {e}")
return None
async def search_memories(self, query: str, memory_types: List[MemoryType] = None,
limit: int = 50) -> List[MemoryEntry]:
"""Search memories based on content and type"""
memories = []
try:
# Build query conditions
conditions = []
params = []
if memory_types:
type_conditions = " OR ".join(["memory_type = ?"] * len(memory_types))
conditions.append(f"({type_conditions})")
params.extend([mt.value for mt in memory_types])
where_clause = " AND ".join(conditions) if conditions else "1=1"
cursor = self.conn.execute(f"""
SELECT * FROM memory_entries
WHERE {where_clause}
ORDER BY importance DESC, last_accessed DESC
LIMIT ?
""", params + [limit])
for row in cursor.fetchall():
memory = MemoryEntry(
memory_id=row[0],
memory_type=MemoryType(row[1]),
content=pickle.loads(row[2]),
timestamp=datetime.fromtimestamp(row[3]),
importance=row[4],
access_count=row[5],
last_accessed=datetime.fromtimestamp(row[6]),
decay_rate=row[7],
tags=set(json.loads(row[8]))
)
# Simple text matching (can be enhanced with vector similarity)
if self._matches_query(memory, query):
memories.append(memory)
except Exception as e:
self.logger.error(f"Error searching memories: {e}")
return sorted(memories, key=lambda m: m.importance, reverse=True)
def _matches_query(self, memory: MemoryEntry, query: str) -> bool:
"""Simple text matching for memory search"""
query_lower = query.lower()
# Search in content
content_str = json.dumps(memory.content).lower()
if query_lower in content_str:
return True
# Search in tags
for tag in memory.tags:
if query_lower in tag.lower():
return True
return False
async def consolidate_memories(self):
"""Consolidate and organize memories"""
try:
# Get all working memories
working_memories = list(self.working_memory)
# Group related memories
memory_groups = self._group_related_memories(working_memories)
# Create consolidated memories
for group in memory_groups:
if len(group) > 1:
consolidated = await self._create_consolidated_memory(group)
await self.store_memory(consolidated)
self.logger.info(f"Consolidated {len(memory_groups)} memory groups")
except Exception as e:
self.logger.error(f"Error consolidating memories: {e}")
def _group_related_memories(self, memories: List[MemoryEntry]) -> List[List[MemoryEntry]]:
"""Group related memories together"""
groups = []
processed = set()
for memory in memories:
if memory.memory_id in processed:
continue
# Find related memories
related = [memory]
for other_memory in memories:
if (other_memory.memory_id != memory.memory_id and
other_memory.memory_id not in processed):
# Simple relatedness check (can be enhanced)
if self._are_memories_related(memory, other_memory):
related.append(other_memory)
processed.add(other_memory.memory_id)
if related:
groups.append(related)
for mem in related:
processed.add(mem.memory_id)
return groups
def _are_memories_related(self, mem1: MemoryEntry, mem2: MemoryEntry) -> bool:
"""Check if two memories are related"""
# Check temporal proximity
time_diff = abs((mem1.timestamp - mem2.timestamp).total_seconds())
if time_diff < 3600: # Within 1 hour
return True
# Check tag overlap
tag_overlap = len(mem1.tags.intersection(mem2.tags))
if tag_overlap > 0:
return True
# Check content similarity (simple approach)
content1 = json.dumps(mem1.content).lower()
content2 = json.dumps(mem2.content).lower()
# Simple word overlap
words1 = set(content1.split())
words2 = set(content2.split())
overlap_ratio = len(words1.intersection(words2)) / max(len(words1), len(words2))
return overlap_ratio > 0.3
async def _create_consolidated_memory(self, memories: List[MemoryEntry]) -> MemoryEntry:
"""Create a consolidated memory from related memories"""
# Combine content
consolidated_content = {
"type": "consolidated",
"source_memories": [mem.memory_id for mem in memories],
"combined_content": [mem.content for mem in memories],
"themes": self._extract_themes(memories)
}
# Calculate importance
importance = max(mem.importance for mem in memories)
# Combine tags
all_tags = set()
for mem in memories:
all_tags.update(mem.tags)
all_tags.add("consolidated")
return MemoryEntry(
memory_type=MemoryType.SEMANTIC,
content=consolidated_content,
importance=importance,
tags=all_tags
)
def _extract_themes(self, memories: List[MemoryEntry]) -> List[str]:
"""Extract common themes from memories"""
# Simple theme extraction (can be enhanced with NLP)
all_text = " ".join([
json.dumps(mem.content) for mem in memories
]).lower()
# Common cybersecurity themes
themes = []
security_themes = [
"vulnerability", "threat", "attack", "exploit", "malware",
"phishing", "social engineering", "network security", "encryption",
"authentication", "authorization", "firewall", "intrusion"
]
for theme in security_themes:
if theme in all_text:
themes.append(theme)
return themes
def _start_background_processes(self):
"""Start background memory management processes"""
def consolidation_worker():
while True:
try:
time.sleep(300) # Every 5 minutes
asyncio.run(self.consolidate_memories())
except Exception as e:
self.logger.error(f"Consolidation error: {e}")
def decay_worker():
while True:
try:
time.sleep(600) # Every 10 minutes
self._apply_memory_decay()
except Exception as e:
self.logger.error(f"Decay error: {e}")
# Start background threads
self.consolidation_thread = threading.Thread(target=consolidation_worker, daemon=True)
self.decay_thread = threading.Thread(target=decay_worker, daemon=True)
self.consolidation_thread.start()
self.decay_thread.start()
self.logger.info("Background memory processes started")
def _apply_memory_decay(self):
"""Apply decay to memories over time"""
try:
cursor = self.conn.execute("""
SELECT memory_id, importance, last_accessed, decay_rate
FROM memory_entries
""")
updates = []
current_time = datetime.now().timestamp()
for row in cursor.fetchall():
memory_id, importance, last_accessed, decay_rate = row
# Calculate time since last access
time_since_access = current_time - last_accessed
# Apply decay (exponential decay)
decay_factor = np.exp(-decay_rate * time_since_access / 86400) # Days
new_importance = importance * decay_factor
# Minimum importance threshold
if new_importance < 0.01:
new_importance = 0.01
updates.append((new_importance, memory_id))
# Batch update
self.conn.executemany("""
UPDATE memory_entries SET importance = ? WHERE memory_id = ?
""", updates)
self.conn.commit()
self.logger.debug(f"Applied decay to {len(updates)} memories")
except Exception as e:
self.logger.error(f"Error applying memory decay: {e}")
class AdvancedReasoningEngine:
"""Advanced reasoning engine with multiple reasoning types"""
def __init__(self, memory_manager: PersistentMemoryManager):
self.memory_manager = memory_manager
self.logger = logging.getLogger("reasoning_engine")
# Reasoning components
self.inference_rules = self._load_inference_rules()
self.reasoning_strategies = {
ReasoningType.DEDUCTIVE: self._deductive_reasoning,
ReasoningType.INDUCTIVE: self._inductive_reasoning,
ReasoningType.ABDUCTIVE: self._abductive_reasoning,
ReasoningType.ANALOGICAL: self._analogical_reasoning,
ReasoningType.CAUSAL: self._causal_reasoning,
ReasoningType.STRATEGIC: self._strategic_reasoning,
ReasoningType.COUNTERFACTUAL: self._counterfactual_reasoning,
ReasoningType.META_COGNITIVE: self._meta_cognitive_reasoning
}
# Active reasoning chains
self.active_chains = {}
def _load_inference_rules(self) -> Dict[str, Dict[str, Any]]:
"""Load inference rules for different reasoning types"""
return {
"modus_ponens": {
"pattern": "If P then Q, P is true",
"conclusion": "Q is true",
"confidence_base": 0.9
},
"modus_tollens": {
"pattern": "If P then Q, Q is false",
"conclusion": "P is false",
"confidence_base": 0.85
},
"hypothetical_syllogism": {
"pattern": "If P then Q, If Q then R",
"conclusion": "If P then R",
"confidence_base": 0.8
},
"disjunctive_syllogism": {
"pattern": "P or Q, not P",
"conclusion": "Q",
"confidence_base": 0.8
},
"causal_inference": {
"pattern": "Event A precedes Event B, correlation observed",
"conclusion": "A may cause B",
"confidence_base": 0.6
}
}
async def start_reasoning_chain(self, topic: str, goal: str,
reasoning_type: ReasoningType = ReasoningType.DEDUCTIVE) -> str:
"""Start a new reasoning chain"""
chain = ReasoningChain(
topic=topic,
goal=goal,
metadata={"reasoning_type": reasoning_type.value}
)
self.active_chains[chain.chain_id] = chain
# Store in memory
memory_entry = MemoryEntry(
memory_type=MemoryType.PROCEDURAL,
content={
"type": "reasoning_chain_start",
"chain_id": chain.chain_id,
"topic": topic,
"goal": goal,
"reasoning_type": reasoning_type.value
},
importance=0.7,
tags={"reasoning", "chain_start", reasoning_type.value}
)
await self.memory_manager.store_memory(memory_entry)
self.logger.info(f"Started reasoning chain: {chain.chain_id}")
return chain.chain_id
async def add_reasoning_step(self, chain_id: str, premise: str,
inference_rule: str = "", evidence: List[str] = None) -> str:
"""Add a step to an active reasoning chain"""
if chain_id not in self.active_chains:
self.logger.error(f"Reasoning chain {chain_id} not found")
return None
chain = self.active_chains[chain_id]
evidence = evidence or []
# Determine reasoning type from chain metadata
reasoning_type = ReasoningType(chain.metadata.get("reasoning_type", "deductive"))
# Apply reasoning strategy
reasoning_func = self.reasoning_strategies.get(reasoning_type, self._deductive_reasoning)
conclusion, confidence = await reasoning_func(premise, inference_rule, evidence, chain)
# Create reasoning step
step = ReasoningStep(
reasoning_type=reasoning_type,
premise=premise,
inference_rule=inference_rule,
conclusion=conclusion,
confidence=confidence,
evidence=evidence,
dependencies=[s.step_id for s in chain.steps[-3:]] # Last 3 steps
)
chain.steps.append(step)
# Store step in memory
memory_entry = MemoryEntry(
memory_type=MemoryType.PROCEDURAL,
content={
"type": "reasoning_step",
"chain_id": chain_id,
"step_id": step.step_id,
"premise": premise,
"conclusion": conclusion,
"confidence": confidence,
"inference_rule": inference_rule
},
importance=confidence,
tags={"reasoning", "step", reasoning_type.value}
)
await self.memory_manager.store_memory(memory_entry)
self.logger.debug(f"Added reasoning step to chain {chain_id}")
return step.step_id
async def _deductive_reasoning(self, premise: str, inference_rule: str,
evidence: List[str], chain: ReasoningChain) -> Tuple[str, float]:
"""Apply deductive reasoning"""
# Look up inference rule
if inference_rule in self.inference_rules:
rule = self.inference_rules[inference_rule]
base_confidence = rule["confidence_base"]
# Apply rule logic (simplified)
if "modus_ponens" in inference_rule.lower():
conclusion = f"Therefore, the consequent follows from the premise: {premise}"
confidence = base_confidence
else:
conclusion = f"Following {inference_rule}: {premise}"
confidence = base_confidence * 0.8
else:
# Default deductive reasoning
conclusion = f"Based on logical deduction from: {premise}"
confidence = 0.7
# Adjust confidence based on evidence
if evidence:
confidence = min(confidence + len(evidence) * 0.05, 0.95)
return conclusion, confidence
async def _inductive_reasoning(self, premise: str, inference_rule: str,
evidence: List[str], chain: ReasoningChain) -> Tuple[str, float]:
"""Apply inductive reasoning"""
# Inductive reasoning builds general conclusions from specific observations
pattern_strength = len(evidence) / max(len(chain.steps) + 1, 1)
conclusion = f"Based on observed pattern in {len(evidence)} cases: {premise}"
confidence = min(0.3 + pattern_strength * 0.4, 0.8) # Inductive reasoning is less certain
return conclusion, confidence
async def _abductive_reasoning(self, premise: str, inference_rule: str,
evidence: List[str], chain: ReasoningChain) -> Tuple[str, float]:
"""Apply abductive reasoning (inference to best explanation)"""
# Abductive reasoning finds the best explanation for observations
explanation_quality = len(evidence) * 0.1
conclusion = f"Best explanation for '{premise}' given available evidence"
confidence = min(0.5 + explanation_quality, 0.75) # Moderate confidence
return conclusion, confidence
async def _analogical_reasoning(self, premise: str, inference_rule: str,
evidence: List[str], chain: ReasoningChain) -> Tuple[str, float]:
"""Apply analogical reasoning"""
# Search for similar past experiences in memory
similar_memories = await self.memory_manager.search_memories(
premise, [MemoryType.EPISODIC], limit=5
)
if similar_memories:
analogy_strength = len(similar_memories) * 0.15
conclusion = f"By analogy to {len(similar_memories)} similar cases: {premise}"
confidence = min(0.4 + analogy_strength, 0.7)
else:
conclusion = f"No strong analogies found for: {premise}"
confidence = 0.3
return conclusion, confidence
async def _causal_reasoning(self, premise: str, inference_rule: str,
evidence: List[str], chain: ReasoningChain) -> Tuple[str, float]:
"""Apply causal reasoning"""
# Look for temporal and correlational patterns
causal_indicators = ["caused by", "resulted in", "led to", "triggered"]
causal_strength = sum(1 for indicator in causal_indicators if indicator in premise.lower())
temporal_evidence = len([e for e in evidence if "time" in e.lower() or "sequence" in e.lower()])
conclusion = f"Causal relationship identified: {premise}"
confidence = min(0.4 + (causal_strength * 0.1) + (temporal_evidence * 0.1), 0.8)
return conclusion, confidence
async def _strategic_reasoning(self, premise: str, inference_rule: str,
evidence: List[str], chain: ReasoningChain) -> Tuple[str, float]:
"""Apply strategic reasoning for long-term planning"""
# Strategic reasoning considers multiple steps and long-term goals
strategic_depth = len(chain.steps)
goal_alignment = 0.8 if chain.goal.lower() in premise.lower() else 0.5
conclusion = f"Strategic implication: {premise} aligns with long-term objectives"
confidence = min(goal_alignment + (strategic_depth * 0.05), 0.85)
return conclusion, confidence
async def _counterfactual_reasoning(self, premise: str, inference_rule: str,
evidence: List[str], chain: ReasoningChain) -> Tuple[str, float]:
"""Apply counterfactual reasoning (what-if scenarios)"""
# Counterfactual reasoning explores alternative scenarios
scenario_plausibility = 0.6 # Default plausibility
if "what if" in premise.lower() or "if not" in premise.lower():
scenario_plausibility += 0.1
conclusion = f"Counterfactual analysis: {premise} would lead to alternative outcomes"
confidence = min(scenario_plausibility, 0.7) # Inherently speculative
return conclusion, confidence
async def _meta_cognitive_reasoning(self, premise: str, inference_rule: str,
evidence: List[str], chain: ReasoningChain) -> Tuple[str, float]:
"""Apply meta-cognitive reasoning (reasoning about reasoning)"""
# Meta-cognitive reasoning evaluates the reasoning process itself
reasoning_quality = sum(step.confidence for step in chain.steps) / max(len(chain.steps), 1)
conclusion = f"Meta-analysis of reasoning quality: {reasoning_quality:.2f} average confidence"
confidence = reasoning_quality
return conclusion, confidence
async def complete_reasoning_chain(self, chain_id: str) -> Optional[ReasoningChain]:
"""Complete a reasoning chain and store results"""
if chain_id not in self.active_chains:
self.logger.error(f"Reasoning chain {chain_id} not found")
return None
chain = self.active_chains[chain_id]
chain.end_time = datetime.now()
# Generate final conclusion
if chain.steps:
# Combine conclusions from all steps
step_conclusions = [step.conclusion for step in chain.steps]
chain.conclusion = f"Final reasoning conclusion: {' → '.join(step_conclusions[-3:])}"
# Calculate overall confidence
confidences = [step.confidence for step in chain.steps]
chain.confidence = sum(confidences) / len(confidences) if confidences else 0.0
chain.success = chain.confidence > 0.5
else:
chain.conclusion = "No reasoning steps completed"
chain.success = False
# Store completed chain in database
try:
self.memory_manager.conn.execute("""
INSERT OR REPLACE INTO reasoning_chains
(chain_id, topic, goal, steps, conclusion, confidence,
start_time, end_time, success, metadata)
VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?)
""", (
chain.chain_id,
chain.topic,
chain.goal,
pickle.dumps(chain.steps),
chain.conclusion,
chain.confidence,
chain.start_time.timestamp(),
chain.end_time.timestamp(),
chain.success,
pickle.dumps(chain.metadata)
))
self.memory_manager.conn.commit()
# Store in episodic memory
memory_entry = MemoryEntry(
memory_type=MemoryType.EPISODIC,
content={
"type": "completed_reasoning_chain",
"chain_id": chain.chain_id,
"topic": chain.topic,
"conclusion": chain.conclusion,
"success": chain.success,
"duration": (chain.end_time - chain.start_time).total_seconds()
},
importance=chain.confidence,
tags={"reasoning", "completed", chain.metadata.get("reasoning_type", "unknown")}
)
await self.memory_manager.store_memory(memory_entry)
# Remove from active chains
del self.active_chains[chain_id]
self.logger.info(f"Completed reasoning chain: {chain_id}")
return chain
except Exception as e:
self.logger.error(f"Error completing reasoning chain: {e}")
return None
class StrategicPlanningEngine:
"""Long-term strategic planning and goal decomposition"""
def __init__(self, memory_manager: PersistentMemoryManager, reasoning_engine: AdvancedReasoningEngine):
self.memory_manager = memory_manager
self.reasoning_engine = reasoning_engine
self.logger = logging.getLogger("strategic_planning")
# Planning templates
self.planning_templates = self._load_planning_templates()
# Active plans
self.active_plans = {}
def _load_planning_templates(self) -> Dict[str, Dict[str, Any]]:
"""Load strategic planning templates"""
return {
"cybersecurity_assessment": {
"phases": [
"reconnaissance",
"vulnerability_analysis",
"threat_modeling",
"risk_assessment",
"mitigation_planning",
"implementation",
"monitoring"
],
"typical_duration": 30, # days
"success_criteria": [
"Complete security posture assessment",
"Identified all critical vulnerabilities",
"Developed mitigation strategies",
"Implemented security controls"
]
},
"penetration_testing": {
"phases": [
"scoping",
"information_gathering",
"threat_modeling",
"vulnerability_assessment",
"exploitation",
"post_exploitation",
"reporting"
],
"typical_duration": 14, # days
"success_criteria": [
"Identified exploitable vulnerabilities",
"Demonstrated business impact",
"Provided remediation recommendations"
]
},
"incident_response": {
"phases": [
"detection",
"analysis",
"containment",
"eradication",
"recovery",
"lessons_learned"
],
"typical_duration": 7, # days
"success_criteria": [
"Contained security incident",
"Minimized business impact",
"Prevented future incidents"
]
}
}
async def create_strategic_plan(self, title: str, primary_goal: str,
template_type: str = "cybersecurity_assessment") -> str:
"""Create a new strategic plan"""
template = self.planning_templates.get(template_type, {})
# Decompose primary goal into sub-goals
sub_goals = await self._decompose_goal(primary_goal, template)
# Create timeline
timeline = self._create_timeline(template, sub_goals)
# Generate milestones
milestones = self._generate_milestones(sub_goals, timeline)
# Assess risks
risk_factors = await self._assess_risks(primary_goal, sub_goals)
# Determine resources
resources_required = self._determine_resources(template, sub_goals)
plan = StrategicPlan(
title=title,
description=f"Strategic plan for {primary_goal}",
primary_goal=primary_goal,
sub_goals=sub_goals,
timeline=timeline,
milestones=milestones,
success_criteria=template.get("success_criteria", []),
risk_factors=risk_factors,
resources_required=resources_required,
current_status="planning"
)
# Store in database
try:
self.memory_manager.conn.execute("""
INSERT INTO strategic_plans
(plan_id, title, description, primary_goal, sub_goals, timeline,
milestones, success_criteria, risk_factors, resources_required,
current_status, progress_percentage, created_at, updated_at)
VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?)
""", (
plan.plan_id,
plan.title,
plan.description,
plan.primary_goal,
pickle.dumps(plan.sub_goals),
pickle.dumps(plan.timeline),
pickle.dumps(plan.milestones),
pickle.dumps(plan.success_criteria),
pickle.dumps(plan.risk_factors),
pickle.dumps(plan.resources_required),
plan.current_status,
plan.progress_percentage,
plan.created_at.timestamp(),
plan.updated_at.timestamp()
))
self.memory_manager.conn.commit()
# Add to active plans
self.active_plans[plan.plan_id] = plan
# Store in episodic memory
memory_entry = MemoryEntry(
memory_type=MemoryType.STRATEGIC,
content={
"type": "strategic_plan_created",
"plan_id": plan.plan_id,
"title": title,
"primary_goal": primary_goal,
"sub_goals_count": len(sub_goals)
},
importance=0.8,
tags={"strategic_planning", "plan_created", template_type}
)
await self.memory_manager.store_memory(memory_entry)
self.logger.info(f"Created strategic plan: {plan.plan_id}")
return plan.plan_id
except Exception as e:
self.logger.error(f"Error creating strategic plan: {e}")
return None
async def _decompose_goal(self, primary_goal: str, template: Dict[str, Any]) -> List[str]:
"""Decompose primary goal into actionable sub-goals"""
# Start reasoning chain for goal decomposition
chain_id = await self.reasoning_engine.start_reasoning_chain(
topic=f"Goal Decomposition: {primary_goal}",
goal="Break down primary goal into actionable sub-goals",
reasoning_type=ReasoningType.STRATEGIC
)
sub_goals = []
# Use template phases if available
if "phases" in template:
for phase in template["phases"]:
sub_goal = f"Complete {phase} phase for {primary_goal}"
sub_goals.append(sub_goal)
# Add reasoning step
await self.reasoning_engine.add_reasoning_step(
chain_id,
f"Phase {phase} is essential for achieving {primary_goal}",
"strategic_decomposition"
)
else:
# Generic decomposition
generic_phases = [
"planning and preparation",
"implementation and execution",
"monitoring and evaluation",
"optimization and improvement"
]
for phase in generic_phases:
sub_goal = f"Complete {phase} for {primary_goal}"
sub_goals.append(sub_goal)
# Complete reasoning chain
await self.reasoning_engine.complete_reasoning_chain(chain_id)
return sub_goals
def _create_timeline(self, template: Dict[str, Any], sub_goals: List[str]) -> Dict[str, datetime]:
"""Create timeline for strategic plan"""
timeline = {}
start_date = datetime.now()
# Total duration from template or estimate
total_duration = template.get("typical_duration", len(sub_goals) * 3) # days
duration_per_goal = total_duration / len(sub_goals) if sub_goals else 1
current_date = start_date
for i, sub_goal in enumerate(sub_goals):
timeline[f"sub_goal_{i}_start"] = current_date
timeline[f"sub_goal_{i}_end"] = current_date + timedelta(days=duration_per_goal)
current_date = timeline[f"sub_goal_{i}_end"]
timeline["plan_start"] = start_date
timeline["plan_end"] = current_date
return timeline
def _generate_milestones(self, sub_goals: List[str], timeline: Dict[str, datetime]) -> List[Dict[str, Any]]:
"""Generate milestones for strategic plan"""
milestones = []
for i, sub_goal in enumerate(sub_goals):
milestone = {
"milestone_id": str(uuid.uuid4()),
"title": f"Milestone {i+1}: {sub_goal}",
"description": f"Complete sub-goal: {sub_goal}",
"target_date": timeline.get(f"sub_goal_{i}_end", datetime.now()),
"success_criteria": [f"Successfully complete {sub_goal}"],
"status": "pending",
"progress_percentage": 0.0
}
milestones.append(milestone)
return milestones
async def _assess_risks(self, primary_goal: str, sub_goals: List[str]) -> List[str]:
"""Assess potential risks for the strategic plan"""
# Start reasoning chain for risk assessment
chain_id = await self.reasoning_engine.start_reasoning_chain(
topic=f"Risk Assessment: {primary_goal}",
goal="Identify potential risks and mitigation strategies",
reasoning_type=ReasoningType.STRATEGIC
)
# Common cybersecurity risks
common_risks = [
"Technical complexity may exceed available expertise",
"Timeline constraints may impact quality",
"Resource availability may be limited",
"External dependencies may cause delays",
"Changing requirements may affect scope",
"Security vulnerabilities may be discovered during implementation",
"Stakeholder availability may be limited"
]
# Assess relevance of each risk
relevant_risks = []
for risk in common_risks:
# Add reasoning step for each risk
await self.reasoning_engine.add_reasoning_step(
chain_id,
f"Risk consideration: {risk}",
"risk_assessment"
)
relevant_risks.append(risk)
# Complete reasoning chain
await self.reasoning_engine.complete_reasoning_chain(chain_id)
return relevant_risks
def _determine_resources(self, template: Dict[str, Any], sub_goals: List[str]) -> List[str]:
"""Determine required resources for strategic plan"""
# Common resources for cybersecurity plans
base_resources = [
"Cybersecurity expertise",
"Technical infrastructure access",
"Documentation and reporting tools",
"Communication and collaboration platforms"
]
# Template-specific resources
if "resources" in template:
base_resources.extend(template["resources"])
# Add resources based on sub-goals
specialized_resources = []
for sub_goal in sub_goals:
if "vulnerability" in sub_goal.lower():
specialized_resources.append("Vulnerability scanning tools")
elif "penetration" in sub_goal.lower():
specialized_resources.append("Penetration testing tools")
elif "monitoring" in sub_goal.lower():
specialized_resources.append("Security monitoring platforms")
return list(set(base_resources + specialized_resources))
async def update_plan_progress(self, plan_id: str, milestone_id: str = None,
progress_percentage: float = None, status: str = None) -> bool:
"""Update progress of strategic plan"""
try:
if plan_id not in self.active_plans:
# Load from database
plan = await self._load_plan(plan_id)
if not plan:
self.logger.error(f"Plan {plan_id} not found")
return False
self.active_plans[plan_id] = plan
plan = self.active_plans[plan_id]
# Update milestone if specified
if milestone_id:
for milestone in plan.milestones:
if milestone["milestone_id"] == milestone_id:
if progress_percentage is not None:
milestone["progress_percentage"] = progress_percentage
if status:
milestone["status"] = status
break
# Update overall plan progress
if progress_percentage is not None:
plan.progress_percentage = progress_percentage
if status:
plan.current_status = status
plan.updated_at = datetime.now()
# Update database
self.memory_manager.conn.execute("""
UPDATE strategic_plans
SET milestones = ?, progress_percentage = ?,
current_status = ?, updated_at = ?
WHERE plan_id = ?
""", (
pickle.dumps(plan.milestones),
plan.progress_percentage,
plan.current_status,
plan.updated_at.timestamp(),
plan_id
))
self.memory_manager.conn.commit()
# Store progress update in memory
memory_entry = MemoryEntry(
memory_type=MemoryType.EPISODIC,
content={
"type": "plan_progress_update",
"plan_id": plan_id,
"milestone_id": milestone_id,
"progress_percentage": progress_percentage,
"status": status
},
importance=0.6,
tags={"strategic_planning", "progress_update"}
)
await self.memory_manager.store_memory(memory_entry)
self.logger.info(f"Updated plan progress: {plan_id}")
return True
except Exception as e:
self.logger.error(f"Error updating plan progress: {e}")
return False
async def _load_plan(self, plan_id: str) -> Optional[StrategicPlan]:
"""Load strategic plan from database"""
try:
cursor = self.memory_manager.conn.execute("""
SELECT * FROM strategic_plans WHERE plan_id = ?
""", (plan_id,))
row = cursor.fetchone()
if row:
return StrategicPlan(
plan_id=row[0],
title=row[1],
description=row[2],
primary_goal=row[3],
sub_goals=pickle.loads(row[4]),
timeline=pickle.loads(row[5]),
milestones=pickle.loads(row[6]),
success_criteria=pickle.loads(row[7]),
risk_factors=pickle.loads(row[8]),
resources_required=pickle.loads(row[9]),
current_status=row[10],
progress_percentage=row[11],
created_at=datetime.fromtimestamp(row[12]),
updated_at=datetime.fromtimestamp(row[13])
)
except Exception as e:
self.logger.error(f"Error loading plan {plan_id}: {e}")
return None
# Integration class that brings everything together
class PersistentCognitiveSystem:
"""Main system that integrates persistent memory, reasoning, and strategic planning"""
def __init__(self, db_path: str = "data/cognitive_system.db"):
# Initialize components
self.memory_manager = PersistentMemoryManager(db_path)
self.reasoning_engine = AdvancedReasoningEngine(self.memory_manager)
self.strategic_planner = StrategicPlanningEngine(self.memory_manager, self.reasoning_engine)
self.logger = logging.getLogger("persistent_cognitive_system")
self.logger.info("Persistent cognitive system initialized")
async def process_complex_scenario(self, scenario: Dict[str, Any]) -> Dict[str, Any]:
"""Process a complex cybersecurity scenario using all cognitive capabilities"""
scenario_id = str(uuid.uuid4())
self.logger.info(f"Processing complex scenario: {scenario_id}")
results = {
"scenario_id": scenario_id,
"timestamp": datetime.now().isoformat(),
"results": {}
}
try:
# Step 1: Store scenario in memory
scenario_memory = MemoryEntry(
memory_type=MemoryType.EPISODIC,
content=scenario,
importance=0.8,
tags={"scenario", "complex", "cybersecurity"}
)
memory_id = await self.memory_manager.store_memory(scenario_memory)
results["results"]["memory_id"] = memory_id
# Step 2: Start strategic planning if it's a long-term objective
if scenario.get("type") == "strategic" or scenario.get("requires_planning", False):
plan_id = await self.strategic_planner.create_strategic_plan(
title=scenario.get("title", f"Scenario {scenario_id}"),
primary_goal=scenario.get("objective", "Complete cybersecurity scenario"),
template_type=scenario.get("template", "cybersecurity_assessment")
)
results["results"]["plan_id"] = plan_id
# Step 3: Apply reasoning to understand the scenario
reasoning_types = scenario.get("reasoning_types", [ReasoningType.DEDUCTIVE])
reasoning_results = {}
for reasoning_type in reasoning_types:
chain_id = await self.reasoning_engine.start_reasoning_chain(
topic=f"Scenario Analysis: {scenario.get('title', scenario_id)}",
goal="Analyze and understand the cybersecurity scenario",
reasoning_type=reasoning_type
)
# Add reasoning steps based on scenario details
for detail in scenario.get("details", []):
await self.reasoning_engine.add_reasoning_step(
chain_id,
detail,
"scenario_analysis",
scenario.get("evidence", [])
)
# Complete reasoning
chain = await self.reasoning_engine.complete_reasoning_chain(chain_id)
reasoning_results[reasoning_type.value] = {
"chain_id": chain_id,
"conclusion": chain.conclusion if chain else "Failed to complete",
"confidence": chain.confidence if chain else 0.0
}
results["results"]["reasoning"] = reasoning_results
# Step 4: Generate recommendations
recommendations = await self._generate_recommendations(scenario, reasoning_results)
results["results"]["recommendations"] = recommendations
# Step 5: Update long-term memory with insights
insight_memory = MemoryEntry(
memory_type=MemoryType.SEMANTIC,
content={
"type": "scenario_insight",
"scenario_id": scenario_id,
"key_learnings": recommendations,
"confidence_scores": {k: v["confidence"] for k, v in reasoning_results.items()}
},
importance=0.7,
tags={"insight", "learning", "cybersecurity"}
)
await self.memory_manager.store_memory(insight_memory)
results["status"] = "success"
self.logger.info(f"Successfully processed scenario: {scenario_id}")
except Exception as e:
results["status"] = "error"
results["error"] = str(e)
self.logger.error(f"Error processing scenario {scenario_id}: {e}")
return results
async def _generate_recommendations(self, scenario: Dict[str, Any],
reasoning_results: Dict[str, Any]) -> List[str]:
"""Generate actionable recommendations based on scenario analysis"""
recommendations = []
# Base recommendations based on scenario type
scenario_type = scenario.get("type", "general")
if scenario_type == "vulnerability_assessment":
recommendations.extend([
"Conduct comprehensive vulnerability scan",
"Prioritize critical vulnerabilities for immediate remediation",
"Implement security patches and updates",
"Establish regular vulnerability monitoring"
])
elif scenario_type == "incident_response":
recommendations.extend([
"Immediately contain the security incident",
"Preserve forensic evidence",
"Assess scope and impact of the incident",
"Implement recovery procedures",
"Conduct post-incident analysis"
])
elif scenario_type == "penetration_testing":
recommendations.extend([
"Define clear scope and objectives",
"Follow structured testing methodology",
"Document all findings and evidence",
"Provide actionable remediation guidance"
])
else:
recommendations.extend([
"Assess current security posture",
"Identify key risk areas",
"Develop mitigation strategies",
"Implement monitoring and detection"
])
# Add reasoning-based recommendations
for reasoning_type, results in reasoning_results.items():
if results["confidence"] > 0.7:
recommendations.append(f"High confidence in {reasoning_type} analysis suggests prioritizing related actions")
# Search for similar past experiences
similar_memories = await self.memory_manager.search_memories(
scenario.get("title", ""), [MemoryType.EPISODIC], limit=3
)
if similar_memories:
recommendations.append(f"Apply lessons learned from {len(similar_memories)} similar past scenarios")
return recommendations[:10] # Limit to top 10 recommendations
# Factory function for easy instantiation
def create_persistent_cognitive_system(db_path: str = "data/cognitive_system.db") -> PersistentCognitiveSystem:
"""Create and initialize the persistent cognitive system"""
return PersistentCognitiveSystem(db_path)
# Main execution for testing
if __name__ == "__main__":
import asyncio
# Configure logging
logging.basicConfig(
level=logging.INFO,
format='%(asctime)s - %(name)s - %(levelname)s - %(message)s'
)
async def test_system():
"""Test the persistent cognitive system"""
# Create system
system = create_persistent_cognitive_system()
# Test scenario
test_scenario = {
"type": "vulnerability_assessment",
"title": "Web Application Security Assessment",
"objective": "Assess security posture of critical web application",
"details": [
"Web application handles sensitive customer data",
"Application has not been tested in 12 months",
"Recent security incidents in similar applications reported"
],
"evidence": [
"Previous vulnerability scan results",
"Security incident reports from industry",
"Application architecture documentation"
],
"reasoning_types": [ReasoningType.DEDUCTIVE, ReasoningType.CAUSAL],
"requires_planning": True,
"template": "cybersecurity_assessment"
}
# Process scenario
results = await system.process_complex_scenario(test_scenario)
print("=== Persistent Cognitive System Test Results ===")
print(json.dumps(results, indent=2, default=str))
# Test memory search
memories = await system.memory_manager.search_memories("vulnerability", limit=5)
print(f"\n=== Found {len(memories)} memories related to 'vulnerability' ===")
for memory in memories:
print(f"- {memory.memory_id}: {memory.content.get('type', 'Unknown')} (importance: {memory.importance:.2f})")
# Run test
asyncio.run(test_system())
|