refactor(memory): add PilotWritePipeline and enrich extraction schema

- Add dedicated PilotWritePipeline (statement → triplet → graph_build → layer-1 dedup, no Neo4j write)
- Add type_description/predicate_description fields across entity and triplet models, Cypher queries, and graph builders
- Refactor data_pruning with LRU cache and snapshot support; skip assistant chunks in extraction
- Remove strict Predicate enum whitelist; support statement_text alias in legacy extractor
- Wire PipelineSnapshot through preprocessing and emotion extraction for debug tracing
- Add PILOT_RUN_USE_REFACTORED_PIPELINE env toggle for pipeline selection

api/app/core/config.py

@@ -272,6 +272,12 @@ class Settings:
     
     MEMORY_OUTPUT_DIR: str = os.getenv("MEMORY_OUTPUT_DIR", "logs/memory-output")
     MEMORY_CONFIG_DIR: str = os.getenv("MEMORY_CONFIG_DIR", "app/core/memory")
+    # Pilot run pipeline switch:
+    # true  -> use refactored PilotWritePipeline
+    # false -> use legacy ExtractionOrchestrator pipeline
+    PILOT_RUN_USE_REFACTORED_PIPELINE: bool = (
+        os.getenv("PILOT_RUN_USE_REFACTORED_PIPELINE", "true").lower() == "true"
+    )
 
     # Tool Management Configuration
     TOOL_CONFIG_DIR: str = os.getenv("TOOL_CONFIG_DIR", "app/core/tools")