dao-ai 0.1.2__py3-none-any.whl → 0.1.3__py3-none-any.whl

dao_ai/cli.py

@@ -715,7 +715,15 @@ def run_databricks_command(
     target: Optional[str] = None,
     dry_run: bool = False,
 ) -> None:
-    """Execute a databricks CLI command with optional profile and target."""
+    """Execute a databricks CLI command with optional profile and target.
+    
+    Args:
+        command: The databricks CLI command to execute (e.g., ["bundle", "deploy"])
+        profile: Optional Databricks CLI profile name
+        config: Optional path to the configuration file
+        target: Optional bundle target name
+        dry_run: If True, print the command without executing
+    """
     config_path = Path(config) if config else None
 
     if config_path and not config_path.exists():
@@ -737,15 +745,17 @@ def run_databricks_command(
         logger.debug(f"Using app-specific target: {target}")
 
     # Build databricks command (no -c flag needed, uses databricks.yaml in current dir)
+    # Note: --profile is a global flag, but --target is a subcommand flag for 'bundle'
     cmd = ["databricks"]
     if profile:
         cmd.extend(["--profile", profile])
 
+    cmd.extend(command)
+
+    # --target must come after the bundle subcommand (it's a subcommand-specific flag)
     if target:
         cmd.extend(["--target", target])
 
-    cmd.extend(command)
-
     # Add config_path variable for notebooks
     if config_path and app_config:
         # Calculate relative path from notebooks directory to config file

dao_ai/config.py

@@ -1009,27 +1009,92 @@ class VolumePathModel(BaseModel, HasFullName):
 
 
 class VectorStoreModel(IsDatabricksResource):
+    """
+    Configuration model for a Databricks Vector Search store.
+
+    Supports two modes:
+    1. **Use Existing Index**: Provide only `index` (fully qualified name).
+       Used for querying an existing vector search index at runtime.
+    2. **Provisioning Mode**: Provide `source_table` + `embedding_source_column`.
+       Used for creating a new vector search index.
+
+    Examples:
+        Minimal configuration (use existing index):
+        ```yaml
+        vector_stores:
+          products_search:
+            index:
+              name: catalog.schema.my_index
+        ```
+
+        Full provisioning configuration:
+        ```yaml
+        vector_stores:
+          products_search:
+            source_table:
+              schema: *my_schema
+              name: products
+            embedding_source_column: description
+            endpoint:
+              name: my_endpoint
+        ```
+    """
+
     model_config = ConfigDict(use_enum_values=True, extra="forbid")
-    embedding_model: Optional[LLMModel] = None
+
+    # RUNTIME: Only index is truly required for querying existing indexes
     index: Optional[IndexModel] = None
+
+    # PROVISIONING ONLY: Required when creating a new index
+    source_table: Optional[TableModel] = None
+    embedding_source_column: Optional[str] = None
+    embedding_model: Optional[LLMModel] = None
     endpoint: Optional[VectorSearchEndpoint] = None
-    source_table: TableModel
+
+    # OPTIONAL: For both modes
     source_path: Optional[VolumePathModel] = None
     checkpoint_path: Optional[VolumePathModel] = None
     primary_key: Optional[str] = None
     columns: Optional[list[str]] = Field(default_factory=list)
     doc_uri: Optional[str] = None
-    embedding_source_column: str
+
+    @model_validator(mode="after")
+    def validate_configuration_mode(self) -> Self:
+        """
+        Validate that configuration is valid for either:
+        - Use existing mode: index is provided
+        - Provisioning mode: source_table + embedding_source_column provided
+        """
+        has_index = self.index is not None
+        has_source_table = self.source_table is not None
+        has_embedding_col = self.embedding_source_column is not None
+
+        # Must have at least index OR source_table
+        if not has_index and not has_source_table:
+            raise ValueError(
+                "Either 'index' (for existing indexes) or 'source_table' "
+                "(for provisioning) must be provided"
+            )
+
+        # If provisioning mode, need embedding_source_column
+        if has_source_table and not has_embedding_col:
+            raise ValueError(
+                "embedding_source_column is required when source_table is provided (provisioning mode)"
+            )
+
+        return self
 
     @model_validator(mode="after")
     def set_default_embedding_model(self) -> Self:
-        if not self.embedding_model:
+        # Only set default embedding model in provisioning mode
+        if self.source_table is not None and not self.embedding_model:
             self.embedding_model = LLMModel(name="databricks-gte-large-en")
         return self
 
     @model_validator(mode="after")
     def set_default_primary_key(self) -> Self:
-        if self.primary_key is None:
+        # Only auto-discover primary key in provisioning mode
+        if self.primary_key is None and self.source_table is not None:
             from dao_ai.providers.databricks import DatabricksProvider
 
             provider: DatabricksProvider = DatabricksProvider()
@@ -1050,14 +1115,16 @@ class VectorStoreModel(IsDatabricksResource):
 
     @model_validator(mode="after")
     def set_default_index(self) -> Self:
-        if self.index is None:
+        # Only generate index from source_table in provisioning mode
+        if self.index is None and self.source_table is not None:
             name: str = f"{self.source_table.name}_index"
             self.index = IndexModel(schema=self.source_table.schema_model, name=name)
         return self
 
     @model_validator(mode="after")
     def set_default_endpoint(self) -> Self:
-        if self.endpoint is None:
+        # Only find/create endpoint in provisioning mode
+        if self.endpoint is None and self.source_table is not None:
             from dao_ai.providers.databricks import (
                 DatabricksProvider,
                 with_available_indexes,
@@ -1549,11 +1616,13 @@ class RerankParametersModel(BaseModel):
             top_n: 5  # Return top 5 after reranking
         ```
 
-    Available models (from fastest to most accurate):
-    - "ms-marco-TinyBERT-L-2-v2" (fastest, smallest)
-    - "ms-marco-MiniLM-L-6-v2"
-    - "ms-marco-MiniLM-L-12-v2" (default, good balance)
-    - "rank-T5-flan" (most accurate, slower)
+    Available models (see https://github.com/PrithivirajDamodaran/FlashRank):
+    - "ms-marco-TinyBERT-L-2-v2" (~4MB, fastest)
+    - "ms-marco-MiniLM-L-12-v2" (~34MB, best cross-encoder, default)
+    - "rank-T5-flan" (~110MB, best non cross-encoder)
+    - "ms-marco-MultiBERT-L-12" (~150MB, multilingual 100+ languages)
+    - "ce-esci-MiniLM-L12-v2" (e-commerce optimized, Amazon ESCI)
+    - "miniReranker_arabic_v1" (Arabic language)
     """
 
     model_config = ConfigDict(use_enum_values=True, extra="forbid")

dao_ai/providers/databricks.py

@@ -625,6 +625,22 @@ class DatabricksProvider(ServiceProvider):
                 df.write.mode("overwrite").saveAsTable(table)
 
     def create_vector_store(self, vector_store: VectorStoreModel) -> None:
+        # Validate that this is a provisioning-mode config
+        if vector_store.source_table is None:
+            raise ValueError(
+                "Cannot create vector store: source_table is required for provisioning. "
+                "This VectorStoreModel appears to be configured for 'use existing index' mode. "
+                "To provision a new vector store, provide source_table and embedding_source_column."
+            )
+        if vector_store.embedding_source_column is None:
+            raise ValueError(
+                "Cannot create vector store: embedding_source_column is required for provisioning."
+            )
+        if vector_store.endpoint is None:
+            raise ValueError(
+                "Cannot create vector store: endpoint is required for provisioning."
+            )
+
         if not endpoint_exists(self.vsc, vector_store.endpoint.name):
             self.vsc.create_endpoint_and_wait(
                 name=vector_store.endpoint.name,

{dao_ai-0.1.2.dist-info → dao_ai-0.1.3.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: dao-ai
-Version: 0.1.2
+Version: 0.1.3
 Summary: DAO AI: A modular, multi-agent orchestration framework for complex AI workflows. Supports agent handoff, tool integration, and dynamic configuration via YAML.
 Project-URL: Homepage, https://github.com/natefleming/dao-ai
 Project-URL: Documentation, https://natefleming.github.io/dao-ai
@@ -79,7 +79,7 @@ Description-Content-Type: text/markdown
 
 # DAO: Declarative Agent Orchestration
 
-[![Version](https://img.shields.io/badge/version-0.1.0-blue.svg)](CHANGELOG.md)
+[![Version](https://img.shields.io/badge/version-0.1.2-blue.svg)](CHANGELOG.md)
 [![Python](https://img.shields.io/badge/python-3.11+-green.svg)](https://www.python.org/)
 [![License](https://img.shields.io/badge/license-MIT-blue.svg)](LICENSE)
 

{dao_ai-0.1.2.dist-info → dao_ai-0.1.3.dist-info}/RECORD

@@ -1,8 +1,8 @@
 dao_ai/__init__.py,sha256=18P98ExEgUaJ1Byw440Ct1ty59v6nxyWtc5S6Uq2m9Q,1062
 dao_ai/agent_as_code.py,sha256=xIlLDpPVfmDVzLvbdY_V_CrC4Jvj2ItCWJ-NzdrszTo,538
 dao_ai/catalog.py,sha256=sPZpHTD3lPx4EZUtIWeQV7VQM89WJ6YH__wluk1v2lE,4947
-dao_ai/cli.py,sha256=azkzrG0RCKyb_euXsmiLCMXWtz7n-8KMNG-45pufqUg,32655
-dao_ai/config.py,sha256=o8BYHB-9R3PVWP0rtGAVrge3CHQqqnIZgfUl_awNPTU,117044
+dao_ai/cli.py,sha256=1TS4I6LQq6ixxd0JQA1bt3DHRizcSiynZkSBcZA3-XE,33149
+dao_ai/config.py,sha256=FmXXwFMWqSxl2ofxPIL2ZLRbx7kx-u6VjFx4wdQnFVk,119755
 dao_ai/graph.py,sha256=1-uQlo7iXZQTT3uU8aYu0N5rnhw5_g_2YLwVsAs6M-U,1119
 dao_ai/logging.py,sha256=lYy4BmucCHvwW7aI3YQkQXKJtMvtTnPDu9Hnd7_O4oc,1556
 dao_ai/messages.py,sha256=4ZBzO4iFdktGSLrmhHzFjzMIt2tpaL-aQLHOQJysGnY,6959
@@ -42,7 +42,7 @@ dao_ai/orchestration/supervisor.py,sha256=0aQLL1a3oBjG9LYTuV1y73rB0mqkIkax7MwtZ7
 dao_ai/orchestration/swarm.py,sha256=8tp1eGmsQqqWpaDcjPoJckddPWohZdmmN0RGRJ_xzOA,9198
 dao_ai/providers/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 dao_ai/providers/base.py,sha256=-fjKypCOk28h6vioPfMj9YZSw_3Kcbi2nMuAyY7vX9k,1383
-dao_ai/providers/databricks.py,sha256=4IBV9_staE6vMci4hUjHWYHc8iS87mM8lT6LX0xh8Mw,61040
+dao_ai/providers/databricks.py,sha256=63jUkvFvdQoJGUlSEJgCsq6hPtj6tviE-AW_TPJvuvc,61850
 dao_ai/tools/__init__.py,sha256=SRd7W2DOCXKbWWy8lclRtJiCskz7SDAm94qaFF47urQ,1664
 dao_ai/tools/agent.py,sha256=plIWALywRjaDSnot13nYehBsrHRpBUpsVZakoGeajOE,1858
 dao_ai/tools/core.py,sha256=bRIN3BZhRQX8-Kpu3HPomliodyskCqjxynQmYbk6Vjs,3783
@@ -57,8 +57,8 @@ dao_ai/tools/sql.py,sha256=tKd1gjpLuKdQDyfmyYYtMiNRHDW6MGRbdEVaeqyB8Ok,7632
 dao_ai/tools/time.py,sha256=tufJniwivq29y0LIffbgeBTIDE6VgrLpmVf8Qr90qjw,9224
 dao_ai/tools/unity_catalog.py,sha256=AjQfW7bvV8NurqDLIyntYRv2eJuTwNdbvex1L5CRjOk,15534
 dao_ai/tools/vector_search.py,sha256=oe2uBwl2TfeJIXPpwiS6Rmz7wcHczSxNyqS9P3hE6co,14542
-dao_ai-0.1.2.dist-info/METADATA,sha256=hNysdpnl7dp_T1gC5tJQ4YHwSLoeg2UJY0aHg4-gvYc,15615
-dao_ai-0.1.2.dist-info/WHEEL,sha256=WLgqFyCfm_KASv4WHyYy0P3pM_m7J5L9k2skdKLirC8,87
-dao_ai-0.1.2.dist-info/entry_points.txt,sha256=Xa-UFyc6gWGwMqMJOt06ZOog2vAfygV_DSwg1AiP46g,43
-dao_ai-0.1.2.dist-info/licenses/LICENSE,sha256=YZt3W32LtPYruuvHE9lGk2bw6ZPMMJD8yLrjgHybyz4,1069
-dao_ai-0.1.2.dist-info/RECORD,,
+dao_ai-0.1.3.dist-info/METADATA,sha256=0WeeUk6J2uQRdk9Hm1FkCk75RG8psnh3WrSt4A8MfNY,15615
+dao_ai-0.1.3.dist-info/WHEEL,sha256=WLgqFyCfm_KASv4WHyYy0P3pM_m7J5L9k2skdKLirC8,87
+dao_ai-0.1.3.dist-info/entry_points.txt,sha256=Xa-UFyc6gWGwMqMJOt06ZOog2vAfygV_DSwg1AiP46g,43
+dao_ai-0.1.3.dist-info/licenses/LICENSE,sha256=YZt3W32LtPYruuvHE9lGk2bw6ZPMMJD8yLrjgHybyz4,1069
+dao_ai-0.1.3.dist-info/RECORD,,