File size: 2,508 Bytes
82a3b34 c624cb8 82a3b34 c624cb8 82a3b34 c624cb8 2a7f7e3 c624cb8 82a3b34 c624cb8 82a3b34 c624cb8 82a3b34 | 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 | """
AegisLM Agents Package
Multi-agent adversarial LLM evaluation framework.
This package contains:
- Attacker agent: Generates adversarial prompts
- Defender agent: Detects and blocks malicious inputs/outputs
- Judge agent: Evaluates model robustness
- Mutation agent: Mutates prompts for diversity
"""
from agents.attacker import (
AttackEngine,
AttackRequest,
AttackResponse,
AttackLog,
BaseAttackStrategy,
PromptInjectionStrategy,
JailbreakStrategy,
BiasTriggerStrategy,
ContextPoisoningStrategy,
RoleConfusionStrategy,
ChainingStrategy,
get_attack_engine,
ATTACK_REGISTRY,
get_attack_strategy,
list_attack_strategies,
register_attack_strategy,
initialize_registry,
)
from agents.defender import (
BaseDefenseStrategy,
InjectionDefenseStrategy,
RoleOverrideDefenseStrategy,
ToxicityDefenseStrategy,
SemanticPolicyDefenseStrategy,
DefenderEngine,
DefenderRequest,
DefenderResponse,
DefenderLog,
DefenderConfig,
get_defender_engine,
DEFENSE_REGISTRY,
DefenseStrategy,
DefaultDefenseStrategy,
PromptSanitizationStrategy,
OutputFilteringStrategy,
get_defense_strategy,
list_defense_strategies,
register_defense_strategy,
initialize_registry as init_defense_registry,
)
__version__ = "0.1.0"
__all__ = [
# Attacker
"AttackEngine",
"AttackRequest",
"AttackResponse",
"AttackLog",
"BaseAttackStrategy",
"PromptInjectionStrategy",
"JailbreakStrategy",
"BiasTriggerStrategy",
"ContextPoisoningStrategy",
"RoleConfusionStrategy",
"ChainingStrategy",
"get_attack_engine",
"ATTACK_REGISTRY",
"get_attack_strategy",
"list_attack_strategies",
"register_attack_strategy",
"initialize_registry",
# Defender Base
"BaseDefenseStrategy",
"InjectionDefenseStrategy",
"RoleOverrideDefenseStrategy",
"ToxicityDefenseStrategy",
"SemanticPolicyDefenseStrategy",
# Defender
"DefenderEngine",
"DefenderRequest",
"DefenderResponse",
"DefenderLog",
"DefenderConfig",
"get_defender_engine",
"DEFENSE_REGISTRY",
"DefenseStrategy",
"DefaultDefenseStrategy",
"PromptSanitizationStrategy",
"OutputFilteringStrategy",
"get_defense_strategy",
"list_defense_strategies",
"register_defense_strategy",
"init_defense_registry",
]
|