Google Professional-Agentic-Architect Value Pack (Frequently Bought Together)

You are developing an autonomous financial research agent using Agent Development Kit (ADK). The agent has access to a registry containing more than 50 distinct financial analysis tools. When you bind all 50 tools to the agent at once, the system prompt consumes too many tokens, causing high latency, increased costs, and frequent LLM hallucinations in which the agent selects the wrong tool schema. You need to improve the agent's tool-calling accuracy and reduce token consumption while ensuring all 50 tools remain available for the workflow.
What should you do?
Choose two

• A. Combine a subset of all 50 tools into a single ADK ToolChain object to compress the tool schemas into a single system prompt description.
• B. Implement an ADK semantic tool retriever to query a vector index and only fetch the top-K most relevant tools based on the user's current prompt.
• C. Increase the model's temperature parameter to allow the agent to find the correct tool schema from memory.
• D. Dynamically bind only a retrieved subset of tools to the agent's execution context before initiating the reasoning loop.
• E. Configure an ADK asynchronous execution hook to run tools in parallel and return only the valid data payloads to the agent.

You are configuring custom coding agents with Antigravity for an enterprise engineering team. You find that the agent frequently attempts to run unauthorized terminal commands, despite instructions in the project's markdown context files that specify command boundaries. You need to enforce strict, deterministic guardrails that validate shell command arguments before tool execution.
What should you do?

• A. Configure an Antigravity PreToolUse hook in a .agents/hooks.json file to run a validation script to inspect command line flags from stdin.
• B. Configure an Antigravity PreToolUse hook in a .agents/hooks.json file to run a script that sends the command string to a remote LLM API for a semantic safety review.
• C. Define a specialized Antigravity subagent in the .agents/subagents/directory configured with a review prompt to evaluate and approve terminal commands prior to delegating execution.
• D. Update the Antigravity workspace rule activation mode to 'Always On' in the .agents/rules/directory to ensure system instructions for command boundary checking are prepended to every conversation turn.

You are deploying a custom Antigravity CLI plugin to enforce enterprise data loss prevention compliance across developer workstations. You must configure an event hook that intercepts tool calls immediately before file execution, automatically redacts sensitive API keys from generated code diffs in transit, and stops tool execution if sanitization fails.
What should you do?

• A. Define a Transform hook in a hooks.json file bound to the pre-tool execution event, and implement a handler function that mutates the payload parameter stream before returning it to the execution context.
• B. Define a postToolUse hook in a hooks.json file bound to the post-tool execution event, and implement an asynchronous handler function that cleans the modified file on disk after execution completes.
• C. Define an Inspect hook in a hooks.json file bound to the pre-tool execution event, and implement a handler function that reads the payload and rewrites the parameters before returning a success status.
• D. Define a Decide hook in a hooks.json file bound to the pre-tool execution event, and implement a handler function that evaluates policy rules and modifies sensitive text strings in-place.

You are deploying an autonomous coding agent on Google Kubernetes Engine (GKE) to run user-requested Python scripts. To isolate untrusted code execution, you enabled GKE Agent Sandbox with gVisor to intercept container system calls in user space. During a security audit, you find that an untrusted Python script generated by a prompt injection attack successfully fetched IAM identity tokens by sending an HTTP request to the link-local instance metadata server at 169.254.169.254.
You need to prevent sandboxed code from exfiltrating metadata tokens while maintaining container kernel isolation.
What should you do?

• A. Apply a Kubernetes NetworkPolicy to the sandbox Pod that blocks ingress traffic to 169.254.169.254.
• B. Reconfigure the gVisor Sentry process to disable the execve system call for child Python processes.
• C. Apply domain restrictions to the sandbox Pod.
• D. Apply a Kubernetes NetworkPolicy to the sandbox Pod that blocks egress traffic to 169.254.169.254.

You are configuring enterprise agents that must dynamically discover and invoke capabilities across managed databases, third-party SaaS tools, and Google Cloud services. Your architecture hosts custom MCP servers on Cloud Run alongside pre-built Google Cloud MCP servers. Your operations team requires a centralized catalog to govern access controls, register JSON tool specifications, and manage IAM identities across all remote HTTP transport endpoints.
You need to maximize operational efficiency and maintain strict governance by establishing a single point of discovery and access control for both native and custom agent capabilities.
What should you do?

• A. Configure custom MCP servers on Cloud Run, and grant granular IAM roles to extension service accounts to manage agent access controls within defined VPC Service Controls perimeters.
• B. Deploy OpenAPI specifications for custom Cloud Run MCP servers using Cloud Endpoints, and register these endpoints in Service Directory to enable dynamic agent service discovery and endpoint- level IAM security.
• C. Register custom Cloud Run MCP servers as Service resources within Agent Registry, upload their JSON tool specifications, and enable built-in Google Cloud MCP servers to centralize agent capability discovery and IAM access governance.
• D. Store local client configuration files containing remote HTTP endpoints in Secret Manager, and mount them across agent execution environments to handle tool discovery and endpoint authentication.