approved changes, still under test

Author: Zvi Fried

src/mcp_as_a_judge/coding_task_manager.py

@@ -10,7 +10,7 @@
 
 from mcp_as_a_judge.db.conversation_history_service import ConversationHistoryService
 from mcp_as_a_judge.logging_config import get_logger
-from mcp_as_a_judge.models.task_metadata import TaskMetadata, TaskState
+from mcp_as_a_judge.models.task_metadata import TaskMetadata, TaskSize, TaskState
 
 # Set up logger using custom get_logger function
 logger = get_logger(__name__)
@@ -23,6 +23,7 @@ async def create_new_coding_task(
     user_requirements: str,
     tags: list[str],
     conversation_service: ConversationHistoryService,
+    task_size: TaskSize,
 ) -> TaskMetadata:
     """
     Create a new coding task with auto-generated task_id.
@@ -34,10 +35,12 @@ async def create_new_coding_task(
         user_requirements: Initial requirements
         tags: Task tags
         conversation_service: Conversation service
+        task_size: Task size classification (REQUIRED)
 
     Returns:
         New TaskMetadata instance
     """
+
     logger.info(f"📝 Creating new coding task: {task_title}")
 
     # Create new TaskMetadata with auto-generated UUID
@@ -46,6 +49,7 @@ async def create_new_coding_task(
         description=task_description,
         user_requirements=user_requirements,
         state=TaskState.CREATED,  # Default state for new tasks
+        task_size=task_size,
         tags=tags,
     )
 

src/mcp_as_a_judge/logging_config.py

@@ -89,7 +89,6 @@ def configure_application_loggers(level: int = logging.INFO) -> None:
     app_loggers = [
         "mcp_as_a_judge.server",
         "mcp_as_a_judge.server_helpers",
-        "mcp_as_a_judge.conversation_history_service",
         "mcp_as_a_judge.db.conversation_history_service",
         "mcp_as_a_judge.db.providers.in_memory",
         "mcp_as_a_judge.db.providers.sqlite_provider",

src/mcp_as_a_judge/models.py

@@ -103,6 +103,74 @@ class ResearchValidationResponse(BaseModel):
     )
 
 
+class ResearchComplexityFactors(BaseModel):
+    """Analysis factors for determining research complexity."""
+
+    domain_specialization: str = Field(
+        description="Level of domain specialization: 'general', 'specialized', 'highly_specialized'"
+    )
+    technology_maturity: str = Field(
+        description="Maturity of required technologies: 'established', 'emerging', 'cutting_edge'"
+    )
+    integration_scope: str = Field(
+        description="Scope of system integration: 'isolated', 'moderate', 'system_wide'"
+    )
+    existing_solutions: str = Field(
+        description="Availability of existing solutions: 'abundant', 'limited', 'scarce'"
+    )
+    risk_level: str = Field(
+        description="Implementation risk level: 'low', 'medium', 'high'"
+    )
+
+
+class ResearchRequirementsAnalysis(BaseModel):
+    """LLM analysis of research requirements for a task."""
+
+    expected_url_count: int = Field(
+        description="Recommended number of research URLs for optimal coverage",
+        ge=0,
+        le=10
+    )
+    minimum_url_count: int = Field(
+        description="Minimum acceptable number of URLs for basic adequacy",
+        ge=0,
+        le=5
+    )
+    reasoning: str = Field(
+        description="Detailed explanation of why these URL counts are appropriate"
+    )
+    complexity_factors: ResearchComplexityFactors = Field(
+        description="Breakdown of complexity analysis factors"
+    )
+    quality_requirements: list[str] = Field(
+        default_factory=list,
+        description="Specific requirements for research source quality and types"
+    )
+
+
+class URLValidationResult(BaseModel):
+    """Result of validating provided URLs against dynamic requirements."""
+
+    adequate: bool = Field(
+        description="Whether the provided URLs meet the dynamic requirements"
+    )
+    provided_count: int = Field(
+        description="Number of URLs actually provided"
+    )
+    expected_count: int = Field(
+        description="Expected number of URLs based on analysis"
+    )
+    minimum_count: int = Field(
+        description="Minimum acceptable number of URLs"
+    )
+    feedback: str = Field(
+        description="Detailed feedback about URL adequacy and suggestions"
+    )
+    meets_quality_standards: bool = Field(
+        description="Whether URLs meet quality requirements beyond just count"
+    )
+
+
 # Database models for conversation history
 # ConversationRecord is now defined in db/interface.py using SQLModel
 # DatabaseConfig is now defined in constants.py
@@ -181,6 +249,20 @@ class JudgeCodingPlanUserVars(BaseModel):
         description="Strategies to mitigate identified risks"
     )
 
+    # Dynamic URL requirements fields - NEW
+    expected_url_count: int = Field(
+        default=0,
+        description="LLM-determined expected number of research URLs for this task"
+    )
+    minimum_url_count: int = Field(
+        default=0,
+        description="LLM-determined minimum acceptable URL count"
+    )
+    url_requirement_reasoning: str = Field(
+        default="",
+        description="LLM explanation of why specific URL count is needed"
+    )
+
 
 class JudgeCodeChangeSystemVars(BaseModel):
     """Variables for judge_code_change system prompt."""
@@ -223,7 +305,7 @@ class ResearchValidationUserVars(BaseModel):
     research: str = Field(description="Research findings to be validated")
     research_urls: list[str] = Field(
         default_factory=list,
-        description="URLs from MANDATORY online research - minimum 3 URLs required",
+        description="URLs from online research - count determined dynamically based on task complexity",
     )
     context: str = Field(description="Additional context about the research validation")
     conversation_history: list = Field(
@@ -241,14 +323,22 @@ class WorkflowGuidanceSystemVars(BaseModel):
 
 
 class WorkflowGuidanceUserVars(BaseModel):
-    """Variables for build_workflow user prompt."""
+    """Variables for workflow_guidance user prompt."""
 
+    task_id: str = Field(description="Task ID")
+    task_title: str = Field(description="Task title")
     task_description: str = Field(description="Description of the development task")
-    context: str = Field(description="Additional context about the task")
-    conversation_history: list = Field(
-        default_factory=list,
-        description="Previous conversation history as JSON array with timestamps",
-    )
+    user_requirements: str = Field(description="User requirements for the task")
+    current_state: str = Field(description="Current task state")
+    state_description: str = Field(description="Description of current state")
+    current_operation: str = Field(description="Current operation being performed")
+    task_size: str = Field(description="Task size classification")
+    task_size_definitions: str = Field(description="Task size definitions")
+    state_transitions: str = Field(description="Valid state transitions")
+    tool_descriptions: str = Field(description="Available tool descriptions")
+    conversation_context: str = Field(description="Conversation history context")
+    operation_context: str = Field(description="Current operation context")
+    response_schema: str = Field(description="JSON schema for the expected response format")
 
 
 class ValidationErrorSystemVars(BaseModel):
@@ -300,6 +390,29 @@ class ElicitationFallbackUserVars(BaseModel):
     )
 
 
+class ResearchRequirementsAnalysisSystemVars(BaseModel):
+    """Variables for research_requirements_analysis system prompt."""
+
+    response_schema: str = Field(
+        description="JSON schema for the expected response format"
+    )
+
+
+class ResearchRequirementsAnalysisUserVars(BaseModel):
+    """Variables for research_requirements_analysis user prompt."""
+
+    task_title: str = Field(description="Title of the coding task")
+    task_description: str = Field(description="Detailed description of the task")
+    user_requirements: str = Field(description="User requirements for the task")
+    research_scope: str = Field(description="Current research scope (none/light/deep)")
+    research_rationale: str = Field(
+        description="Rationale for why research is needed at current scope"
+    )
+    context: str = Field(
+        description="Additional context about the task and project"
+    )
+
+
 class TestingEvaluationSystemVars(BaseModel):
     """Variables for testing evaluation system prompt."""
 

src/mcp_as_a_judge/models/__init__.py

@@ -54,6 +54,8 @@
     "ResearchValidationUserVars",
     "WorkflowGuidanceSystemVars",
     "WorkflowGuidanceUserVars",
+    "DynamicSchemaSystemVars",
+    "DynamicSchemaUserVars",
 ]
 
 # Import additional models from the original models.py file
@@ -83,6 +85,8 @@
         ResearchValidationUserVars = models_py.ResearchValidationUserVars
         WorkflowGuidanceSystemVars = models_py.WorkflowGuidanceSystemVars
         WorkflowGuidanceUserVars = models_py.WorkflowGuidanceUserVars
+        DynamicSchemaSystemVars = models_py.DynamicSchemaSystemVars
+        DynamicSchemaUserVars = models_py.DynamicSchemaUserVars
 
 except Exception:
     # Fallback if models.py doesn't exist or has issues

src/mcp_as_a_judge/models/enhanced_responses.py

@@ -10,11 +10,11 @@
 
 from pydantic import BaseModel, Field
 
-from mcp_as_a_judge.models.task_metadata import TaskMetadata
-from mcp_as_a_judge.workflow import WorkflowGuidance
+from mcp_as_a_judge.models.task_metadata import TaskMetadata, TaskSize
+from mcp_as_a_judge.workflow.workflow_guidance import WorkflowGuidance
 
 
-class JudgeResponseWithTask(BaseModel):
+class JudgeResponse(BaseModel):
     """
     Enhanced JudgeResponse that ALWAYS includes current task metadata and workflow guidance.
     
@@ -35,10 +35,22 @@ class JudgeResponseWithTask(BaseModel):
 
     # Enhanced workflow v3 fields
     current_task_metadata: TaskMetadata = Field(
-        description="ALWAYS current state of task metadata after operation"
+        default_factory=lambda: TaskMetadata(
+            title="Unknown Task",
+            description="No metadata provided",
+            user_requirements="",
+            task_size=TaskSize.M,
+        ),
+        description="ALWAYS current state of task metadata after operation",
     )
     workflow_guidance: WorkflowGuidance = Field(
-        description="LLM-generated next steps and instructions from shared method"
+        default_factory=lambda: WorkflowGuidance(
+            next_tool="raise_obstacle",
+            reasoning="Default guidance: insufficient context",
+            preparation_needed=[],
+            guidance="Provide required parameters and context",
+        ),
+        description="LLM-generated next steps and instructions from shared method",
     )
 
 
@@ -149,10 +161,8 @@ class MissingRequirementsResult(BaseModel):
         description="LLM-generated next steps and instructions for requirements clarification"
     )
 
-
-# Backward compatibility aliases for existing code
-# These can be removed once all tools are migrated to enhanced responses
-JudgeResponse = JudgeResponseWithTask  # Alias for backward compatibility
+# Backward compatibility alias
+JudgeResponseWithTask = JudgeResponse
 
 
 class EnhancedResponseFactory:
@@ -170,9 +180,9 @@ def create_judge_response(
         current_task_metadata: TaskMetadata,
         workflow_guidance: WorkflowGuidance,
         required_improvements: Optional[List[str]] = None,
-    ) -> JudgeResponseWithTask:
+    ) -> JudgeResponse:
         """
-        Create a JudgeResponseWithTask with consistent structure.
+        Create a JudgeResponse with consistent structure.
         
         Args:
             approved: Whether the validation passed
@@ -182,9 +192,9 @@ def create_judge_response(
             required_improvements: Optional list of required improvements
             
         Returns:
-            JudgeResponseWithTask instance
+            JudgeResponse instance
         """
-        return JudgeResponseWithTask(
+        return JudgeResponse(
             approved=approved,
             feedback=feedback,
             required_improvements=required_improvements or [],