feat: add severity field to GeneratedQuestionCard

feat: replace keyword gap analyzer with LLM-based analyze_gaps

refactor: replace GapAnalysisConfig keyword fields with LLM model config

refactor: update test fixtures to use Pydantic models

refactor: update CrewOrchestrator to return ChatTurnResponse

refactor: update BaseOrchestrator and MockChatBackend to return ChatTurnResponse

refactor: type analyze_gaps and extract_decisions with AgentResponse

feat: add PreviewData and ChatTurnResponse pydantic models

refactor: remove duplicate AgentResponse/ChatResult dataclasses from core/types

refactor: convert PipelineResult from dataclass to Pydantic BaseModel

refactor: convert AgentDef from dataclass to Pydantic BaseModel

refactor: add ToolConfig pydantic model, replace tool_config dicts

feat: add memory recall/remember, planning, and collaboration to Flow

refactor: migrate tools to BaseTool subclasses with args_schema

refactor: extract derive_part_name, remove dead code from prompts.py

fix: address code review findings from Flow refactor

refactor: wire AgentDispatchFlow, remove RoutingEngine

feat: add AgentDispatchFlow with router-gated agent dispatch

feat: add agent_flow models, routing, and readiness helpers

test: verify plan/approve preserves notes field

feat: add notes field to DesignPlan and field_agents config

fix: prevent question card loop when agent mentions keywords as context

feat: integrate gap analyzer into orchestrator, update CNC/CAM NOT READY prompts

feat: add gap analyzer with analyze_gaps() and generate_question_cards()

feat: add GapAnalysisConfig to settings with category keywords and agents

feat: add 3MF export and /api/models/{name}.3mf endpoint

feat: add planning phase branching to CrewOrchestrator

feat: add /api/plan/approve and /api/plan/reject endpoints

feat: add DesignPlan model, compute_score(), phase/plan fields to DesignState

feat: add PlanningConfig to settings with threshold, weights, caps

refactor: eliminate all Serializer classes, use Pydantic model_dump() directly

refactor: typed Pydantic config + Pydantic result models, eliminate dict access

test: replace SingleCallOrchestrator tests with CrewOrchestrator tests

feat(tools): add generate_gcode_tool, store last shape on execution

feat(cam): add CAMPlan Pydantic model for structured output

fix: update max_tokens test to match config value (8192)

feat(cam): add CAM agent definition and cam config section

feat(cam): add generate_gcode function with ocp-freecad-cam wrapper

feat(cam): add CAMResult dataclass and serializer

feat: add CrewAI CadQuery tools — execute_cad_tool, validate_cad_tool

feat: complete LLM adapter — enable function calling, add vision support

feat: add BaseOrchestrator ABC

feat: add RoutingEngine class with config-driven keywords

feat: add BackendFactory registry for centralized backend creation

feat: add serializers for execution and validation results

feat: add type system — enums, dataclasses, LLMBackend ABC

feat: add Settings singleton with config.yaml + .env loading

feat: expand keyword routing vocabulary for better agent selection

test: add end-to-end pipeline integration tests

test: add CadQuery executor and CNC validator integration tests