agent chat

config/feature_flags.yaml

@@ -44,7 +44,7 @@ flags:
 
   # Agent Zero - AI agent orchestration framework
   agent_zero:
-    enabled: false
+    enabled: true
     description: "Agent Zero - Autonomous AI agents for task automation"
     type: release
 

ushadow/backend/main.py

@@ -21,12 +21,13 @@
 
 from src.routers import health, wizard, chronicle, auth, feature_flags
 from src.routers import services, deployments, providers, instances, chat
-from src.routers import kubernetes, tailscale, unodes, docker
+from src.routers import kubernetes, tailscale, unodes, docker, agent_zero
 from src.routers import settings as settings_api
 from src.middleware import setup_middleware
 from src.services.unode_manager import init_unode_manager, get_unode_manager
 from src.services.deployment_manager import init_deployment_manager
 from src.services.kubernetes_manager import init_kubernetes_manager
+from src.services.agent_zero import init_agent_zero_service
 from src.services.feature_flags import create_feature_flag_service, set_feature_flag_service
 from src.services.mcp_server import setup_mcp_server
 from src.config.omegaconf_settings import get_settings_store
@@ -141,6 +142,10 @@ def send_telemetry():
     await init_kubernetes_manager(db)
     logger.info("✓ Kubernetes manager initialized")
 
+    # Initialize Agent Zero service
+    await init_agent_zero_service(db)
+    logger.info("✓ Agent Zero service initialized")
+
     # Start background task for stale u-node checking
     stale_check_task = asyncio.create_task(check_stale_unodes_task())
 
@@ -178,6 +183,7 @@ def send_telemetry():
 app.include_router(providers.router, prefix="/api/providers", tags=["providers"])
 app.include_router(instances.router, tags=["instances"])
 app.include_router(chat.router, prefix="/api/chat", tags=["chat"])
+app.include_router(agent_zero.router, prefix="/api/agent-zero", tags=["agent-zero"])
 app.include_router(deployments.router, tags=["deployments"])
 app.include_router(tailscale.router, tags=["tailscale"])
 

ushadow/backend/src/models/agent.py

@@ -0,0 +1,177 @@
+"""
+Agent models for Agent Zero - autonomous AI agents for task automation.
+
+Agents are created from natural language descriptions in chat and can be
+triggered based on conversation context.
+"""
+
+from datetime import datetime
+from enum import Enum
+from typing import Dict, List, Optional, Any
+
+from pydantic import BaseModel, Field
+
+
+class AgentStatus(str, Enum):
+    """Status of an agent."""
+    ACTIVE = "active"          # Agent is active and will respond to triggers
+    INACTIVE = "inactive"      # Agent is paused/disabled
+    DRAFT = "draft"            # Agent is being configured
+
+
+class AgentTrigger(BaseModel):
+    """
+    Defines when an agent should be activated.
+
+    Agents can be triggered by:
+    - Keywords/phrases in the conversation
+    - Explicit invocation by name
+    - Context matching (semantic similarity)
+    """
+    type: str = Field(
+        default="keyword",
+        description="Trigger type: keyword, context, explicit"
+    )
+    keywords: List[str] = Field(
+        default_factory=list,
+        description="Keywords/phrases that trigger the agent"
+    )
+    context_description: Optional[str] = Field(
+        default=None,
+        description="Description of context when agent should activate"
+    )
+    threshold: float = Field(
+        default=0.7,
+        description="Similarity threshold for context matching (0-1)"
+    )
+
+
+class AgentOutput(BaseModel):
+    """
+    Defines how an agent should structure its output.
+    """
+    format: str = Field(
+        default="markdown",
+        description="Output format: markdown, json, plain"
+    )
+    sections: List[str] = Field(
+        default_factory=list,
+        description="Required sections in the output"
+    )
+    include_sources: bool = Field(
+        default=False,
+        description="Whether to include sources/citations"
+    )
+
+
+class Agent(BaseModel):
+    """
+    An autonomous AI agent for task automation.
+
+    Agents are created from natural language descriptions and can be
+    triggered based on conversation context to perform specific tasks.
+    """
+    id: str = Field(..., description="Unique agent ID")
+    name: str = Field(..., description="Agent display name")
+    description: str = Field(..., description="What the agent does")
+
+    # What triggers this agent
+    trigger: AgentTrigger = Field(
+        default_factory=AgentTrigger,
+        description="When this agent should activate"
+    )
+
+    # What the agent does
+    system_prompt: str = Field(
+        default="",
+        description="System prompt for the agent's LLM"
+    )
+    instructions: str = Field(
+        default="",
+        description="Detailed instructions for the agent"
+    )
+
+    # Output configuration
+    output: AgentOutput = Field(
+        default_factory=AgentOutput,
+        description="How the agent should format output"
+    )
+
+    # Status
+    status: AgentStatus = Field(
+        default=AgentStatus.ACTIVE,
+        description="Current agent status"
+    )
+
+    # Metadata
+    created_at: Optional[datetime] = None
+    updated_at: Optional[datetime] = None
+    created_by: Optional[str] = None
+    last_used_at: Optional[datetime] = None
+    use_count: int = Field(default=0, description="Number of times agent was invoked")
+
+    # Tags for organization
+    tags: List[str] = Field(default_factory=list)
+    metadata: Dict[str, Any] = Field(default_factory=dict)
+
+    class Config:
+        use_enum_values = True
+
+
+class AgentCreate(BaseModel):
+    """Request to create a new agent."""
+    name: str = Field(..., min_length=1, max_length=100)
+    description: str = Field(..., min_length=1, max_length=500)
+    trigger: Optional[AgentTrigger] = None
+    system_prompt: Optional[str] = None
+    instructions: Optional[str] = None
+    output: Optional[AgentOutput] = None
+    tags: List[str] = Field(default_factory=list)
+
+
+class AgentUpdate(BaseModel):
+    """Request to update an agent."""
+    name: Optional[str] = None
+    description: Optional[str] = None
+    trigger: Optional[AgentTrigger] = None
+    system_prompt: Optional[str] = None
+    instructions: Optional[str] = None
+    output: Optional[AgentOutput] = None
+    status: Optional[AgentStatus] = None
+    tags: Optional[List[str]] = None
+
+
+class AgentFromChat(BaseModel):
+    """
+    Request to create an agent from natural language in chat.
+
+    The LLM will parse this to extract agent configuration.
+    """
+    user_request: str = Field(..., description="The user's natural language request")
+    conversation_context: Optional[List[Dict[str, str]]] = Field(
+        default=None,
+        description="Previous conversation messages for context"
+    )
+
+
+class AgentInvocation(BaseModel):
+    """
+    Record of an agent being invoked.
+    """
+    id: str = Field(..., description="Unique invocation ID")
+    agent_id: str = Field(..., description="The agent that was invoked")
+    trigger_type: str = Field(..., description="How the agent was triggered")
+    input_context: str = Field(..., description="The input that triggered the agent")
+    output: str = Field(..., description="The agent's response")
+    created_at: datetime = Field(default_factory=datetime.utcnow)
+    user_id: Optional[str] = None
+
+
+class AgentExecuteRequest(BaseModel):
+    """Request to execute an agent with specific input."""
+    agent_id: str = Field(..., description="Agent to execute")
+    input_text: str = Field(..., description="Input for the agent")
+    additional_context: Optional[Dict[str, Any]] = Field(
+        default=None,
+        description="Additional context to provide to the agent"
+    )