added guardrails.py to constrain responses and added validation to config.py

python/ebook_search/config.py

@@ -3,9 +3,9 @@
 from __future__ import annotations
 
 from os import getenv
-from typing import Annotated
+from typing import Annotated, Self
 
-from pydantic import AliasChoices, Field, field_validator
+from pydantic import AliasChoices, Field, field_validator, model_validator
 from pydantic_settings import BaseSettings, NoDecode, SettingsConfigDict
 
 
@@ -82,6 +82,8 @@ class EbookSearchConfig(BaseSettings):
     vector_candidate_multiplier: int = 4
     bm25_candidate_limit: int = 120
     rrf_rank_constant: int = 60
+    min_retrieval_confidence: float = 0.0
+    validate_citations_enabled: bool = True
     bm25_index_dir: str = ".ebook_search_bm25"
     bm25_refresh_delay_seconds: int = 60
 
@@ -99,6 +101,20 @@ class EbookSearchConfig(BaseSettings):
         """Normalize the configured embedding alias to its provider model name."""
         return normalize_embedding_alias(value)
 
+    @model_validator(mode="after")
+    def validate_runtime_consistency(self) -> Self:
+        """Reject configurations that cannot serve the features they enable."""
+        if not self.embedding_base_url.strip():
+            msg = "embedding_base_url must be set"
+            raise ValueError(msg)
+        if self.answer_enabled and (not self.vllm_base_url.strip() or not self.chat_model.strip()):
+            msg = "answer_enabled requires vllm_base_url and chat_model to be set"
+            raise ValueError(msg)
+        if self.rerank.enabled and not self.rerank.base_url.strip():
+            msg = "rerank.enabled requires rerank.base_url to be set"
+            raise ValueError(msg)
+        return self
+
 
 def load_rerank_config() -> RerankConfig:
     """Load reranker config from environment variables."""