The Agent Skills Directory

Cloud Resource Management: The skill utilizes the gcloud CLI and official Google Cloud SDKs to manage resources such as BigQuery datasets, Vertex AI Vector Search collections, and Cloud Run services. These operations are limited to the GCP project configured by the user in design-spec.md and represent standard cloud infrastructure management tasks.
Workspace Initialization: A shell script (bootstrap.sh) is provided to automate the creation of a Python virtual environment and the installation of the skill's dependencies. This is a common development workflow designed to ensure the agent has access to the necessary libraries in a controlled environment.
Dependency Management: The skill specifies well-known and official libraries in its pyproject.toml, such as google-cloud-aiplatform and google-adk. These are hosted on standard package registries and are appropriate for the skill's stated purpose of building Google Cloud-based agents.
Data Ingestion Surface: The skill is designed to process product catalog data from CSV or JSON files. As a Retrieval-Augmented Generation (RAG) agent, it incorporates this data into the AI's context. While this is the intended functionality, it is a standard security practice to ensure that any data files provided to the system are from trusted and verified sources.

retail-product-search