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
"""Response-action plugin base class."""
from __future__ import annotations
from abc import abstractmethod
from typing import Any
from pydantic import BaseModel, Field
from .plugin import AiSOCPlugin, PluginContext
class ActionRequest(BaseModel):
"""Payload sent to a response-action plugin."""
action_id: str = Field(..., description="Registered action identifier")
params: dict[str, Any] = Field(default_factory=dict, description="Action-specific parameters")
dry_run: bool = Field(
False,
description="If True the plugin must validate params and return what would happen without executing",
)
case_id: str | None = Field(None, description="Associated case ID, if any")
playbook_run_id: str | None = Field(None, description="Playbook run that triggered this action")
class ActionResult(BaseModel):
"""Result returned by a response-action plugin."""
action_id: str
success: bool
dry_run: bool = False
summary: str = Field(default="", description="Human-readable one-liner of what happened")
details: dict[str, Any] = Field(default_factory=dict, description="Structured action output")
error: str | None = None
class ActionPlugin(AiSOCPlugin):
"""Base class for response-action plugins."""
@abstractmethod
def supported_actions(self) -> list[str]:
"""Return list of action IDs this plugin handles."""
@abstractmethod
async def execute(self, request: ActionRequest, ctx: PluginContext) -> ActionResult:
"""Execute the action and return a structured result."""