agent-starter-pack 0.1.7__py3-none-any.whl → 0.2.1__py3-none-any.whl

src/base_template/deployment/terraform/storage.tf

@@ -12,8 +12,13 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
+provider "google" {
+  region = var.region
+  user_project_override = true
+}
+
 resource "google_storage_bucket" "bucket_load_test_results" {
-  name                        = "${var.cicd_runner_project_id}-${var.suffix_bucket_name_load_test_results}"
+  name                        = "${var.cicd_runner_project_id}-${var.project_name}-load-test"
   location                    = var.region
   project                     = var.cicd_runner_project_id
   uniform_bucket_level_access = true
@@ -23,7 +28,7 @@ resource "google_storage_bucket" "bucket_load_test_results" {
 
 resource "google_storage_bucket" "logs_data_bucket" {
   for_each                    = toset(local.all_project_ids)
-  name                        = "${each.value}-logs-data"
+  name                        = "${each.value}-${var.project_name}-logs-data"
   location                    = var.region
   project                     = each.value
   uniform_bucket_level_access = true
@@ -31,11 +36,21 @@ resource "google_storage_bucket" "logs_data_bucket" {
 
   depends_on = [resource.google_project_service.cicd_services, resource.google_project_service.shared_services]
 }
+{% if cookiecutter.deployment_target == 'cloud_run' %}
+resource "google_artifact_registry_repository" "repo-artifacts-genai" {
+  location      = var.region
+  repository_id = "${var.project_name}-repo"
+  description   = "Repo for Generative AI applications"
+  format        = "DOCKER"
+  project       = var.cicd_runner_project_id
+  depends_on    = [resource.google_project_service.cicd_services, resource.google_project_service.shared_services]
+}
+{% endif %}
 
-{%- if cookiecutter.data_ingestion %}
+{% if cookiecutter.data_ingestion %}
 resource "google_storage_bucket" "data_ingestion_pipeline_gcs_root" {
   for_each                    = local.deploy_project_ids
-  name                        = "${each.value}-pipeline-artifacts"
+  name                        = "${each.value}-${var.project_name}-rag"
   location                    = var.region
   project                     = each.value
   uniform_bucket_level_access = true
@@ -44,11 +59,13 @@ resource "google_storage_bucket" "data_ingestion_pipeline_gcs_root" {
   depends_on = [resource.google_project_service.cicd_services, resource.google_project_service.shared_services]
 }
 
+{% if cookiecutter.datastore_type == "vertex_ai_search" %}
+
 resource "google_discovery_engine_data_store" "data_store_staging" {
   location                    = var.data_store_region
   project                     = var.staging_project_id
-  data_store_id               = "${var.datastore_name}"
-  display_name                = "${var.datastore_name}"
+  data_store_id               = "${var.project_name}-datastore"
+  display_name                = "${var.project_name}-datastore"
   industry_vertical           = "GENERIC"
   content_config              = "NO_CONTENT"
   solution_types              = ["SOLUTION_TYPE_SEARCH"]
@@ -59,7 +76,7 @@ resource "google_discovery_engine_data_store" "data_store_staging" {
 
 resource "google_discovery_engine_search_engine" "search_engine_staging" {
   project        = var.staging_project_id
-  engine_id      = "${var.search_engine_name}"
+  engine_id      = "${var.project_name}-search"
   collection_id  = "default_collection"
   location       = google_discovery_engine_data_store.data_store_staging.location
   display_name   = "Search Engine App Staging"
@@ -73,8 +90,8 @@ resource "google_discovery_engine_search_engine" "search_engine_staging" {
 resource "google_discovery_engine_data_store" "data_store_prod" {
   location                    = var.data_store_region
   project                     = var.prod_project_id
-  data_store_id               = "${var.datastore_name}"
-  display_name                = "${var.datastore_name}"
+  data_store_id               = "${var.project_name}-datastore"
+  display_name                = "${var.project_name}-datastore"
   industry_vertical           = "GENERIC"
   content_config              = "NO_CONTENT"
   solution_types              = ["SOLUTION_TYPE_SEARCH"]
@@ -85,7 +102,7 @@ resource "google_discovery_engine_data_store" "data_store_prod" {
 
 resource "google_discovery_engine_search_engine" "search_engine_prod" {
   project        = var.prod_project_id
-  engine_id      = "${var.search_engine_name}"
+  engine_id      = "${var.project_name}-search"
   collection_id  = "default_collection"
   location       = google_discovery_engine_data_store.data_store_prod.location
   display_name   = "Search Engine App Prod"
@@ -95,6 +112,101 @@ resource "google_discovery_engine_search_engine" "search_engine_prod" {
   }
   provider = google.prod_billing_override
 }
-{%- endif %}
+{% elif cookiecutter.datastore_type == "vertex_ai_vector_search" %}
+
+resource "google_storage_bucket" "vector_search_data_bucket" {
+  for_each                    = local.deploy_project_ids
+  name                        = "${each.value}-${var.project_name}-vs"
+  location                    = var.region
+  project                     = each.value
+  uniform_bucket_level_access = true
+  force_destroy               = true
+
+  depends_on = [resource.google_project_service.cicd_services, resource.google_project_service.shared_services]
+}
 
+resource "google_vertex_ai_index" "vector_search_index_staging" {
+  project = var.staging_project_id
+  region = var.region
+  display_name = "${var.project_name}-vector-search"
+  description = "vector search index for test"
+  metadata {
+    config {
+      dimensions = var.vector_search_embedding_size
+      shard_size = var.vector_search_shard_size
+      distance_measure_type = "DOT_PRODUCT_DISTANCE"
+      approximate_neighbors_count = var.vector_search_approximate_neighbors_count
+      algorithm_config {
+        tree_ah_config {
+        }
+      }
+    }
+  }
+  index_update_method = "STREAM_UPDATE"
+}
+
+resource "google_vertex_ai_index_endpoint" "vector_search_index_endpoint_staging" {
+  project      = var.staging_project_id
+  region       = var.region
+  display_name = "${var.project_name}-vector-search-endpoint"
+  public_endpoint_enabled = true
+  depends_on = [google_vertex_ai_index.vector_search_index_staging]
+}
+
+resource "google_vertex_ai_index_endpoint_deployed_index" "vector_search_index_deployment_staging" {
+    index_endpoint = google_vertex_ai_index_endpoint.vector_search_index_endpoint_staging.id
+    index = google_vertex_ai_index.vector_search_index_staging.id
+    deployed_index_id = replace("${var.project_name}_deployed_index", "-", "_")
+    depends_on = [
+      google_vertex_ai_index.vector_search_index_staging,
+      google_vertex_ai_index_endpoint.vector_search_index_endpoint_staging
+    ]
+}
+
+resource "google_vertex_ai_index" "vector_search_index_prod" {
+  project = var.prod_project_id
+  region = var.region
+  display_name = "${var.project_name}-vector-search"
+  description = "vector search index for test"
+  metadata {
+    config {
+      dimensions = var.vector_search_embedding_size
+      distance_measure_type = "DOT_PRODUCT_DISTANCE"
+      approximate_neighbors_count = var.vector_search_approximate_neighbors_count
+      shard_size = var.vector_search_shard_size
+      algorithm_config {
+        tree_ah_config {
+        }
+      }
+    }
+  }
+  index_update_method = "STREAM_UPDATE"
+}
+
+resource "google_vertex_ai_index_endpoint" "vector_search_index_endpoint_prod" {
+  project      = var.prod_project_id
+  region       = var.region
+  display_name = "${var.project_name}-vector-search-endpoint"
+  public_endpoint_enabled = true
+  depends_on = [google_vertex_ai_index.vector_search_index_prod]
+}
+
+resource "google_vertex_ai_index_endpoint_deployed_index" "vector_search_index_deployment_prod" {
+    index_endpoint = google_vertex_ai_index_endpoint.vector_search_index_endpoint_prod.id
+    index = google_vertex_ai_index.vector_search_index_prod.id
+    deployed_index_id = replace("${var.project_name}_deployed_index", "-", "_")
+    dedicated_resources {
+      machine_spec {
+        machine_type = var.vector_search_machine_type
+      }
+      min_replica_count = var.vector_search_min_replica_count
+      max_replica_count = var.vector_search_max_replica_count
+    }
+    depends_on = [
+      google_vertex_ai_index.vector_search_index_prod,
+      google_vertex_ai_index_endpoint.vector_search_index_endpoint_prod
+    ]
+}
 
+{% endif %}
+{% endif %}

src/base_template/deployment/terraform/variables.tf

@@ -12,6 +12,12 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
+variable "project_name" {
+  type        = string
+  description = "Project name used as a base for resource naming"
+  default     = "{{ cookiecutter.project_name | replace('_', '-') }}"
+}
+
 variable "prod_project_id" {
   type        = string
   description = "**Production** Google Cloud Project ID for resource deployment."
@@ -43,22 +49,10 @@ variable "repository_name" {
   type        = string
 }
 
-variable "telemetry_bigquery_dataset_id" {
-  type        = string
-  description = "BigQuery dataset ID for telemetry data export."
-  default     = "telemetry_genai_app_sample_sink"
-}
-
-variable "feedback_bigquery_dataset_id" {
-  type        = string
-  description = "BigQuery dataset ID for feedback data export."
-  default     = "feedback_genai_app_sample_sink"
-}
-
 variable "telemetry_logs_filter" {
   type        = string
   description = "Log Sink filter for capturing telemetry data. Captures logs with the `traceloop.association.properties.log_type` attribute set to `tracing`."
-  default     = "jsonPayload.attributes.\"traceloop.association.properties.log_type\"=\"tracing\" jsonPayload.resource.attributes.\"service.name\"=\"Sample Chatbot Application\""
+  default     = "jsonPayload.attributes.\"traceloop.association.properties.log_type\"=\"tracing\" jsonPayload.resource.attributes.\"service.name\"=\"{{cookiecutter.project_name}}\""
 }
 
 variable "feedback_logs_filter" {
@@ -67,48 +61,13 @@ variable "feedback_logs_filter" {
   default     = "jsonPayload.log_type=\"feedback\""
 }
 
-variable "telemetry_sink_name" {
-  type        = string
-  description = "Name of the telemetry data Log Sink."
-  default     = "telemetry_logs_genai_app_sample"
-}
-
-variable "feedback_sink_name" {
-  type        = string
-  description = "Name of the feedback data Log Sink."
-  default     = "feedback_logs_genai_app_sample"
-}
-
-variable "cicd_runner_sa_name" {
-  description = "Service account name to be used for the CICD processes"
-  type        = string
-  default     = "cicd-runner"
-}
-
-variable "suffix_bucket_name_load_test_results" {
-  description = "Suffix Name of the bucket that will be used to store the results of the load test. Prefix will be project id."
-  type        = string
-  default     = "cicd-load-test-results"
-}
-{%- if cookiecutter.deployment_target == 'cloud_run' %}
-variable "cloud_run_app_sa_name" {
-  description = "Service account name to be used for the Cloud Run service"
-  type        = string
-  default     = "{{cookiecutter.project_name}}-cr"
-}
-
-variable "artifact_registry_repo_name" {
-  description = "Name of the Artifact registry repository to be used to push containers"
-  type        = string
-  default     = "genai-containers"
-}
-
+{% if cookiecutter.deployment_target == 'cloud_run' %}
 variable "cloud_run_app_roles" {
   description = "List of roles to assign to the Cloud Run app service account"
-{%- elif cookiecutter.deployment_target == 'agent_engine' %}
+{% elif cookiecutter.deployment_target == 'agent_engine' %}
 variable "agentengine_sa_roles" {
   description = "List of roles to assign to the Agent Engine service account"
-{%- endif %}
+{% endif %}
   type        = list(string)
   default = [
 {%- if cookiecutter.deployment_target == 'cloud_run' %}
@@ -152,12 +111,7 @@ variable "cicd_sa_deployment_required_roles" {
   ]
 }
 
-{%- if cookiecutter.data_ingestion %}
-variable "vertexai_pipeline_sa_name" {
-  description = "Service account name to be used for the Vertex AI service"
-  type        = string
-  default     = "data-ingestion-vertexai-sa"
-}
+{% if cookiecutter.data_ingestion %}
 
 variable "pipeline_cron_schedule" {
   type        = string
@@ -165,12 +119,6 @@ variable "pipeline_cron_schedule" {
   default     = "0 0 * * 0" # Run at 00:00 UTC every Sunday
 }
 
-variable "data_store_region" {
-  type        = string
-  description = "Google Cloud region for resource deployment."
-  default     = "us"
-}
-
 variable "pipelines_roles" {
   description = "List of roles to assign to the Vertex AI Pipelines service account"
   type        = list(string)
@@ -187,16 +135,47 @@ variable "pipelines_roles" {
     "roles/resourcemanager.projectIamAdmin"
   ]
 }
+{% if cookiecutter.datastore_type == "vertex_ai_search" %}
+variable "data_store_region" {
+  type        = string
+  description = "Google Cloud region for resource deployment."
+  default     = "us"
+}
+{% elif cookiecutter.datastore_type == "vertex_ai_vector_search" %}
+variable "vector_search_embedding_size" {
+  type = number
+  description = "The number of dimensions for the embeddings."
+  default = 768
+}
+
+variable "vector_search_approximate_neighbors_count" {
+  type = number
+  description = "The approximate number of neighbors to return."
+  default = 150
+}
+
+variable "vector_search_min_replica_count" {
+  type = number
+  description = "The min replica count for vector search instance"
+  default = 1
+}
 
-variable "datastore_name" {
-  description = "The name of the datastore"
+variable "vector_search_max_replica_count" {
+  type = number
+  description = "The max replica count for vector search instance"
+  default = 1
+}
+
+variable "vector_search_shard_size" {
+  description = "The shard size of the vector search instance"
   type = string
-  default = "sample-datastore"
+  default = "SHARD_SIZE_SMALL"
 }
 
-variable "search_engine_name" {
-  description = "The name of the search engine"
+variable "vector_search_machine_type" {
+  description = "The machine type for the vector search instance"
   type = string
-  default = "sample-search-engine"
+  default = "e2-standard-2"
 }
-{%- endif %}
+{% endif %}
+{% endif %}

src/base_template/deployment/terraform/vars/env.tfvars

@@ -1,3 +1,6 @@
+# Project name used for resource naming
+project_name = "{{ cookiecutter.project_name | replace('_', '-') }}"
+
 # Your Production Google Cloud project id
 prod_project_id = "your-production-project-id"
 
@@ -8,36 +11,28 @@ staging_project_id = "your-staging-project-id"
 cicd_runner_project_id = "your-cicd-project-id"
 
 # Name of the host connection you created in Cloud Build
-host_connection_name = "your-host-connection-name"
+host_connection_name = "git-{{cookiecutter.project_name}}"
 
 # Name of the repository you added to Cloud Build
-repository_name = "your-repository-name"
+repository_name = "repo-{{cookiecutter.project_name}}"
 
 # The Google Cloud region you will use to deploy the infrastructure
 region = "us-central1"
 
-{%- if cookiecutter.deployment_target == 'cloud_run' %}
-cloud_run_app_sa_name = "{{cookiecutter.project_name}}-cr"
-{%- endif %}
-
-telemetry_bigquery_dataset_id = "telemetry_genai_app_sample_sink"
-telemetry_sink_name = "telemetry_logs_genai_app_sample"
 telemetry_logs_filter = "jsonPayload.attributes.\"traceloop.association.properties.log_type\"=\"tracing\" jsonPayload.resource.attributes.\"service.name\"=\"{{cookiecutter.project_name}}\""
 
-feedback_bigquery_dataset_id = "feedback_genai_app_sample_sink"
-feedback_sink_name = "feedback_logs_genai_app_sample"
 feedback_logs_filter = "jsonPayload.log_type=\"feedback\""
 
-cicd_runner_sa_name = "cicd-runner"
-
-suffix_bucket_name_load_test_results = "cicd-load-test-results"
-
 {%- if cookiecutter.data_ingestion %}
-search_engine_name = "sample-search-engine"
-datastore_name = "sample-datastore"
-vertexai_pipeline_sa_name = "vertexai-pipelines-sa"
 pipeline_cron_schedule = "0 0 * * 0"
 
+{%- if cookiecutter.datastore_type == "vertex_ai_search" %}
 #The value can only be one of "global", "us" and "eu".
 data_store_region = "us"
+{%- elif cookiecutter.datastore_type == "vertex_ai_vector_search" %}
+vector_search_shard_size = "SHARD_SIZE_SMALL"
+vector_search_machine_type = "e2-standard-2"
+vector_search_min_replica_count = 1
+vector_search_max_replica_count = 1
+{%- endif %}
 {%- endif %}

src/base_template/pyproject.toml

@@ -18,7 +18,7 @@ dependencies = [
     "fastapi~=0.115.8",
     "uvicorn~=0.34.0"
 {%- elif cookiecutter.deployment_target == 'agent_engine' %}
-    "google-cloud-aiplatform[evaluation,langchain,reasoningengine]~=1.81.0"
+    "google-cloud-aiplatform[evaluation,reasoningengine]~=1.81.0"
 {%- endif %}
 ]
 {% if cookiecutter.deployment_target == 'cloud_run' %}
@@ -87,7 +87,7 @@ follow_imports = "silent"
 ignore_missing_imports = true
 explicit_package_bases = true
 disable_error_code = ["misc", "no-untyped-call", "no-any-return"]
-{% if cookiecutter.agent_name == 'multimodal_live_api' %}
+{% if cookiecutter.agent_name == 'live_api' %}
 exclude = [".venv","./frontend"]
 {% else %}
 exclude = [".venv"]
@@ -95,7 +95,7 @@ exclude = [".venv"]
 
 [tool.codespell]
 ignore-words-list = "rouge"
-{% if cookiecutter.agent_name == 'multimodal_live_api' %}
+{% if cookiecutter.agent_name == 'live_api' %}
 skip = "./locust_env/*,uv.lock,.venv,./frontend,**/*.ipynb"
 {% else %}
 skip = "./locust_env/*,uv.lock,.venv,**/*.ipynb"
@@ -108,6 +108,7 @@ build-backend = "hatchling.build"
 
 [tool.pytest.ini_options]
 pythonpath = "."
+asyncio_default_fixture_loop_scope = "function"
 
 [tool.hatch.build.targets.wheel]
 packages = ["app","frontend"]

src/cli/commands/create.py

@@ -22,13 +22,15 @@ from click.core import ParameterSource
 from rich.console import Console
 from rich.prompt import IntPrompt, Prompt
 
+from ..utils.datastores import DATASTORE_TYPES
 from ..utils.gcp import verify_credentials, verify_vertex_connection
 from ..utils.logging import handle_cli_error
 from ..utils.template import (
     get_available_agents,
     get_template_path,
+    load_template_config,
     process_template,
-    prompt_data_ingestion,
+    prompt_datastore_selection,
     prompt_deployment_target,
 )
 
@@ -46,7 +48,16 @@ console = Console()
     help="Deployment target name",
 )
 @click.option(
-    "--include-data-ingestion", "-i", is_flag=True, help="Include data pipeline"
+    "--include-data-ingestion",
+    "-i",
+    is_flag=True,
+    help="Include data ingestion pipeline in the project",
+)
+@click.option(
+    "--datastore",
+    "-ds",
+    type=click.Choice(DATASTORE_TYPES),
+    help="Type of datastore to use for data ingestion (requires --include-data-ingestion)",
 )
 @click.option("--debug", is_flag=True, help="Enable debug logging")
 @click.option(
@@ -76,6 +87,7 @@ def create(
     agent: str | None,
     deployment_target: str | None,
     include_data_ingestion: bool,
+    datastore: str | None,
     debug: bool,
     output_dir: str | None,
     auto_approve: bool,
@@ -90,6 +102,23 @@ def create(
         console.print(
             "This tool will help you create an end-to-end production-ready AI agent in GCP!\n"
         )
+        # Validate project name
+        if len(project_name) > 26:
+            console.print(
+                f"Error: Project name '{project_name}' exceeds 26 characters. Please use a shorter name.",
+                style="bold red",
+            )
+            return
+
+        # Convert project name to lowercase
+        if any(char.isupper() for char in project_name):
+            original_name = project_name
+            project_name = project_name.lower()
+            console.print(
+                f"Warning: Project name '{original_name}' contains uppercase characters. "
+                f"Converting to lowercase: '{project_name}'",
+                style="bold yellow",
+            )
 
         # Setup debug logging if enabled
         if debug:
@@ -136,6 +165,38 @@ def create(
         if debug:
             logging.debug(f"Selected agent: {agent}")
 
+        # Data ingestion and datastore selection
+        if include_data_ingestion or datastore:
+            # If datastore is specified but include_data_ingestion is not, set it to True
+            include_data_ingestion = True
+
+            # If include_data_ingestion is True but no datastore is specified, prompt for it
+            if not datastore:
+                # Pass a flag to indicate this is from explicit CLI flag
+                datastore = prompt_datastore_selection(final_agent, from_cli_flag=True)
+
+            if debug:
+                logging.debug(f"Data ingestion enabled: {include_data_ingestion}")
+                logging.debug(f"Selected datastore type: {datastore}")
+        else:
+            # Check if the agent requires data ingestion
+            template_path = (
+                pathlib.Path(__file__).parent.parent.parent.parent
+                / "agents"
+                / final_agent
+                / "template"
+            )
+            config = load_template_config(template_path)
+            if config and config.get("settings", {}).get("requires_data_ingestion"):
+                include_data_ingestion = True
+                datastore = prompt_datastore_selection(final_agent)
+
+                if debug:
+                    logging.debug(
+                        f"Data ingestion required by agent: {include_data_ingestion}"
+                    )
+                    logging.debug(f"Selected datastore type: {datastore}")
+
         # Deployment target selection
         final_deployment = (
             deployment_target
@@ -145,12 +206,6 @@ def create(
         if debug:
             logging.debug(f"Selected deployment target: {final_deployment}")
 
-        # Data pipeline selection
-        include_data_ingestion = include_data_ingestion or prompt_data_ingestion(
-            final_agent
-        )
-        if debug:
-            logging.debug(f"Include data pipeline: {include_data_ingestion}")
         # Region confirmation (if not explicitly passed)
         if (
             not auto_approve
@@ -164,7 +219,6 @@ def create(
         logging.debug("Setting up GCP...")
 
         creds_info = {}
-        # Check for uv installation if not skipping checks
         if not skip_checks:
             # Set up GCP environment
             try:
@@ -208,12 +262,16 @@ def create(
             project_name,
             deployment_target=final_deployment,
             include_data_ingestion=include_data_ingestion,
+            datastore=datastore,
             output_dir=destination_dir,
         )
 
+        # Replace region in all files if a different region was specified
+        if region != "us-central1":
+            replace_region_in_files(project_path, region, debug=debug)
+
         project_path = destination_dir / project_name
-        # Determine the correct path to display to the user
-        display_path = str(project_path) if output_dir else project_name
+        cd_path = project_path if output_dir else project_name
 
         if include_data_ingestion:
             project_id = creds_info.get("project", "")
@@ -222,23 +280,25 @@ def create(
                 f"This agent uses a datastore for grounded responses.\n"
                 f"The agent will work without data, but for optimal results:\n"
                 f"1. Set up dev environment:\n"
-                f"   [white italic]`export PROJECT_ID={project_id} && cd {display_path} && make setup-dev-env`[/white italic]\n\n"
+                f"   [white italic]`export PROJECT_ID={project_id} && cd {cd_path} && make setup-dev-env`[/white italic]\n\n"
                 f"   See deployment/README.md for more info\n"
                 f"2. Run the data ingestion pipeline:\n"
-                f"   [white italic]`export PROJECT_ID={project_id} && cd {display_path} && make data-ingestion`[/white italic]\n\n"
+                f"   [white italic]`export PROJECT_ID={project_id} && cd {cd_path} && make data-ingestion`[/white italic]\n\n"
                 f"   See data_ingestion/README.md for more info\n"
                 f"[bold white]=================================[/bold white]\n"
             )
         console.print("\n> 👍 Done. Execute the following command to get started:")
 
-        # Show the command to get started
+        console.print("\n> Success! Your agent project is ready.")
         console.print(
-            f"[bold bright_green]cd {display_path} && make install && make playground[/]"
+            "\n📖 For more information on project structure, usage, and deployment, check out the README:"
+        )
+        console.print(f"   [cyan]cat {cd_path}/README.md[/]")
+        # Determine the correct path to display based on whether output_dir was specified
+        console.print("\n🚀 To get started, run the following command:")
+        console.print(
+            f"   [bold bright_green]cd {cd_path} && make install && make playground[/]"
         )
-
-        # Replace region in all files if a different region was specified
-        if region != "us-central1":
-            replace_region_in_files(project_path, region, debug=debug)
     except Exception:
         if debug:
             logging.exception(