fix: rag tool embeddings config

* fix: ensure config is not flattened, add tests

* chore: refactor inits to model_validator

* chore: refactor rag tool config parsing

* chore: add initial docs

* chore: add additional validation aliases for provider env vars

* chore: add solid docs

* chore: move imports to top

* fix: revert circular import

* fix: lazy import qdrant-client

* fix: allow collection name config

* chore: narrow model names for google

* chore: update additional docs

* chore: add backward compat on model name aliases

* chore: add tests for config changes

docs/en/tools/file-document/pdfsearchtool.mdx

@@ -63,15 +63,15 @@ tool = PDFSearchTool(
             "config": {
                 # Model identifier for the chosen provider. "model" will be auto-mapped to "model_name" internally.
                 "model": "text-embedding-3-small",
-                # Optional: API key. If omitted, the tool will use provider-specific env vars when available
-                # (e.g., OPENAI_API_KEY for provider="openai").
+                # Optional: API key. If omitted, the tool will use provider-specific env vars
+                # (e.g., OPENAI_API_KEY or EMBEDDINGS_OPENAI_API_KEY for OpenAI).
                 # "api_key": "sk-...",
 
                 # Provider-specific examples:
                 # --- Google Generative AI ---
                 # (Set provider="google-generativeai" above)
-                # "model": "models/embedding-001",
-                # "task_type": "retrieval_document",
+                # "model_name": "gemini-embedding-001",
+                # "task_type": "RETRIEVAL_DOCUMENT",
                 # "title": "Embeddings",
 
                 # --- Cohere ---