mirror of
https://github.com/Significant-Gravitas/AutoModerate.git
synced 2026-11-09 04:59:33 -10:00
Bentlybro
dd455d4dd3
Project and moderation rule creation methods now return dictionaries instead of ORM instances to avoid detached instance issues. Updated related usages in dashboard route and default rules configuration to match new return types and argument names. Also added cascade delete-orphan to APIUser's project relationship.
78 lines
2.9 KiB
Python
78 lines
2.9 KiB
Python
"""
|
|
Default moderation rules configuration
|
|
"""
|
|
|
|
DEFAULT_MODERATION_RULES = [
|
|
{
|
|
"name": "Fraud & Impersonation",
|
|
"description": "Content that misrepresents identity, scams users, or spreads fraudulent schemes.",
|
|
"rule_type": "ai_prompt",
|
|
"rule_data": {"prompt": "Content that misrepresents identity, scams users, or spreads fraudulent schemes."},
|
|
"action": "reject",
|
|
"priority": 100,
|
|
"is_active": True,
|
|
},
|
|
{
|
|
"name": "Phishing & Unauthorized Data Collection",
|
|
"description": ("Any attempt to collect user data unlawfully, "
|
|
"including deceptive AI-generated content designed to steal credentials."),
|
|
"rule_type": "ai_prompt",
|
|
"rule_data": {"prompt": ("Any attempt to collect user data unlawfully, "
|
|
"including deceptive AI-generated content designed to steal credentials.")},
|
|
"action": "reject",
|
|
"priority": 100,
|
|
"is_active": True,
|
|
},
|
|
{
|
|
"name": "Misleading AI Content",
|
|
"description": ("AI-generated content that spreads false information, "
|
|
"deepfakes, or impersonates individuals without disclosure."),
|
|
"rule_type": "ai_prompt",
|
|
"rule_data": {"prompt": ("AI-generated content that spreads false information, "
|
|
"deepfakes, or impersonates individuals without disclosure.")},
|
|
"action": "reject",
|
|
"priority": 100,
|
|
"is_active": True,
|
|
},
|
|
{
|
|
"name": "Illegal Content",
|
|
"description": ("Content that violates applicable laws, "
|
|
"including terrorism, child exploitation, and financial crimes."),
|
|
"rule_type": "ai_prompt",
|
|
"rule_data": {"prompt": ("Content that violates applicable laws, "
|
|
"including terrorism, child exploitation, and financial crimes.")},
|
|
"action": "reject",
|
|
"priority": 100,
|
|
"is_active": True,
|
|
},
|
|
{
|
|
"name": "Spam & Unsolicited Promotions",
|
|
"description": "Unwanted advertising, excessive marketing, and pyramid schemes.",
|
|
"rule_type": "ai_prompt",
|
|
"rule_data": {"prompt": "Unwanted advertising, excessive marketing, and pyramid schemes."},
|
|
"action": "reject",
|
|
"priority": 100,
|
|
"is_active": True,
|
|
},
|
|
]
|
|
|
|
|
|
async def create_default_rules(db_service, project_id):
|
|
"""
|
|
Create default moderation rules for a new project
|
|
|
|
Args:
|
|
db_service: Database service instance
|
|
project_id: ID of the project to create rules for
|
|
"""
|
|
for rule in DEFAULT_MODERATION_RULES:
|
|
await db_service.create_moderation_rule(
|
|
project_id=project_id,
|
|
name=rule["name"],
|
|
description=rule["description"],
|
|
rule_type=rule["rule_type"],
|
|
rule_data=rule["rule_data"],
|
|
action=rule["action"],
|
|
priority=rule["priority"]
|
|
)
|