claude-mpm 3.9.11__py3-none-any.whl → 4.0.3__py3-none-any.whl

claude_mpm/models/agent_definition.py

@@ -19,15 +19,16 @@ DESIGN DECISION: Using dataclasses for models because:
 from dataclasses import dataclass, field
 from datetime import datetime
 from enum import Enum
-from typing import Dict, List, Optional, Any
+from typing import Any, Dict, List, Optional
 
 
 class AgentType(str, Enum):
     """Agent type classification.
-    
+
     WHY: Enum ensures only valid agent types are used throughout the system,
     preventing typos and making the code more maintainable.
     """
+
     CORE = "core"
     PROJECT = "project"
     CUSTOM = "custom"
@@ -37,10 +38,11 @@ class AgentType(str, Enum):
 
 class AgentSection(str, Enum):
     """Agent markdown section identifiers.
-    
+
     WHY: Standardizes section names across the codebase, making it easier
     to parse and update specific sections programmatically.
     """
+
     PRIMARY_ROLE = "Primary Role"
     WHEN_TO_USE = "When to Use This Agent"
     CAPABILITIES = "Core Capabilities"
@@ -55,12 +57,13 @@ class AgentSection(str, Enum):
 @dataclass
 class AgentPermissions:
     """Agent authority and permissions.
-    
+
     WHY: Separating permissions into a dedicated class allows for:
     - Clear permission boundaries
     - Easy permission checking and validation
     - Future extension without modifying the main agent definition
     """
+
     exclusive_write_access: List[str] = field(default_factory=list)
     forbidden_operations: List[str] = field(default_factory=list)
     read_access: List[str] = field(default_factory=list)
@@ -69,12 +72,13 @@ class AgentPermissions:
 @dataclass
 class AgentWorkflow:
     """Agent workflow definition.
-    
+
     WHY: Workflows are complex structures that benefit from their own model:
     - Ensures consistent workflow structure
     - Makes workflow validation easier
     - Allows workflow-specific operations
     """
+
     name: str
     trigger: str
     process: List[str]
@@ -85,12 +89,13 @@ class AgentWorkflow:
 @dataclass
 class AgentMetadata:
     """Agent metadata information.
-    
+
     WHY: Metadata is separated from the main definition because:
     - It changes independently of agent behavior
     - It's used for discovery and management, not execution
     - Different services may need different metadata views
     """
+
     type: AgentType
     model_preference: str = "claude-3-sonnet"
     version: str = "1.0.0"
@@ -98,17 +103,17 @@ class AgentMetadata:
     author: Optional[str] = None
     tags: List[str] = field(default_factory=list)
     specializations: List[str] = field(default_factory=list)
-    
+
     def increment_serial_version(self) -> None:
         """Increment the patch version number.
-        
+
         WHY: Automatic version incrementing ensures every change is tracked
         and follows semantic versioning principles.
         """
-        parts = self.version.split('.')
+        parts = self.version.split(".")
         if len(parts) == 3:
             parts[2] = str(int(parts[2]) + 1)
-            self.version = '.'.join(parts)
+            self.version = ".".join(parts)
         else:
             # If version format is unexpected, append .1
             self.version = f"{self.version}.1"
@@ -117,25 +122,26 @@ class AgentMetadata:
 @dataclass
 class AgentDefinition:
     """Complete agent definition.
-    
+
     WHY: This is the main model that represents an agent's complete configuration:
     - Combines all aspects of an agent in one place
     - Provides a clear contract for what constitutes an agent
     - Makes serialization/deserialization straightforward
-    
+
     DESIGN DECISION: Using composition over inheritance:
     - AgentMetadata, AgentPermissions, and AgentWorkflow are separate classes
     - This allows each component to evolve independently
     - Services can work with just the parts they need
     """
+
     # Core identifiers
     name: str
     title: str
     file_path: str
-    
+
     # Metadata
     metadata: AgentMetadata
-    
+
     # Agent behavior definition
     primary_role: str
     when_to_use: Dict[str, List[str]]  # {"select": [...], "do_not_select": [...]}
@@ -146,14 +152,14 @@ class AgentDefinition:
     kpis: List[str]
     dependencies: List[str]
     tools_commands: str
-    
+
     # Raw content for preservation
     raw_content: str = ""
     raw_sections: Dict[str, str] = field(default_factory=dict)
-    
+
     def to_dict(self) -> Dict[str, Any]:
         """Convert to dictionary for API responses.
-        
+
         WHY: Many parts of the system need agent data as dictionaries:
         - JSON serialization for APIs
         - Configuration storage
@@ -167,10 +173,14 @@ class AgentDefinition:
                 "type": self.metadata.type.value,
                 "model_preference": self.metadata.model_preference,
                 "version": self.metadata.version,
-                "last_updated": self.metadata.last_updated.isoformat() if self.metadata.last_updated else None,
+                "last_updated": (
+                    self.metadata.last_updated.isoformat()
+                    if self.metadata.last_updated
+                    else None
+                ),
                 "author": self.metadata.author,
                 "tags": self.metadata.tags,
-                "specializations": self.metadata.specializations
+                "specializations": self.metadata.specializations,
             },
             "primary_role": self.primary_role,
             "when_to_use": self.when_to_use,
@@ -178,19 +188,19 @@ class AgentDefinition:
             "authority": {
                 "exclusive_write_access": self.authority.exclusive_write_access,
                 "forbidden_operations": self.authority.forbidden_operations,
-                "read_access": self.authority.read_access
+                "read_access": self.authority.read_access,
             },
             "workflows": [
                 {
                     "name": w.name,
                     "trigger": w.trigger,
                     "process": w.process,
-                    "output": w.output
+                    "output": w.output,
                 }
                 for w in self.workflows
             ],
             "escalation_triggers": self.escalation_triggers,
             "kpis": self.kpis,
             "dependencies": self.dependencies,
-            "tools_commands": self.tools_commands
-        }
+            "tools_commands": self.tools_commands,
+        }