Open-source AI Security Operations Center: alert fusion, LLM-agent triage, MITRE ATT&CK investigation, and a replayable decision ledger for every agent step. Self-hostable, runs with no API keys, MIT licensed. Ships an MCP server for Claude, Cursor and Continue.
Python agentic-aiai-securityalert-triageblue-teamcybersecuritydetection-engineeringincident-responselanggraphllm-agentsmcpmitre-attackmodel-context-protocolpurple-teamsecurity-operationsself-hostedsiemsoarsocthreat-detectionthreat-intelligence
"""Data-source connector plugin base class."""
from __future__ import annotations
from abc import abstractmethod
from typing import Any, AsyncIterator
from pydantic import BaseModel, Field
from .plugin import AiSOCPlugin, PluginContext
class ConnectorConfig(BaseModel):
"""Configuration for a connector plugin instance."""
connector_id: str = Field(..., description="Unique connector instance identifier")
enabled: bool = True
poll_interval_seconds: int = Field(60, ge=10, description="Polling interval in seconds")
extra: dict[str, Any] = Field(default_factory=dict, description="Connector-specific settings")
class ConnectorPlugin(AiSOCPlugin):
"""Base class for data-source connector plugins.
Connectors are responsible for ingesting events from external systems
(SIEMs, ticketing systems, cloud logs, etc.) and normalising them into
the AiSOC event format.
"""
@abstractmethod
async def test_connection(self, ctx: PluginContext) -> bool:
"""Verify connectivity to the upstream data source. Return True if healthy."""
@abstractmethod
async def fetch_events(
self,
ctx: PluginContext,
since: str | None = None,
) -> AsyncIterator[dict[str, Any]]:
"""Yield normalised events from the upstream source.
Args:
ctx: Plugin context with credentials and config.
since: ISO-8601 timestamp cursor; fetch events after this point.
Yields:
Normalised event dicts conforming to the AiSOC OCSF event schema.
"""