Consolidated Chapter Plan

Detailed Table of Contents

Chapter 0: The PAA Reference Stack & Core Architecture Principles

• 0.1 The End-to-End Reference Architecture Topology (Mermaid Blueprint)
• 0.2 Layer-by-Layer Service Responsibility Breakdown
• 0.3 Core Architecture Principles of the PAA Exam (If Signal X → Eliminate Trap Y → Select Winner Z)

Chapter 1: Low-Code Agentic Workflows, Prompting & Multimodal Data Stores

• Subsection 1.1: Configuring Agentic Workflows and Behavior Using Low-Code Tools (Part I: Building Agents Using Low-Code Tools (~13% of Exam))
• Domain Overview and Reference Architecture
• Core Foundational Primitives & Default Behaviors (1.1)
• Consideration 1.1a — Configuring state-based workflows (pages, transition routes, and event handlers) using gemini enterprise tools (e.g., gemini enterprise agent designer and customer experience agent studio [cx agent studio]): Page Entry Fulfillment Execution Order; Tiered Fallback Handling via Event Handlers; Dynamic Routing via Parameter-Based Condition Routes
• Consideration 1.1b — Creating system instructions and in-console prompt templates (e.g., few-shot and chain-of-thought) to guide agent behavior (e.g., agent designer and cx agent studio): Few-Shot Prompt Formatting in CX Agent Studio; Step-by-Step Reasoning via Chain-of-Thought in Agent Designer; Persona and Role Definition via System Instructions; Dynamic Variable Substitution in Console Prompt Templates
• Multi-Constraint Production Scenarios, Trade-Off Clinic & Pattern-Trigger Cheat Sheet (1.1)
• Subsection 1.2: Connecting Enterprise Data and Unstructured Multimodal Pipelines to Gemini Enterprise (Part I: Building Agents Using Low-Code Tools (~13% of Exam))
• Domain Overview and Reference Architecture
• Core Foundational Primitives & Default Behaviors (1.2)
• Consideration 1.2a — Configuring agents to securely connect and query enterprise proprietary data sources (e.g., gemini enterprise and agent search): End-User Identity and ACL Propagation in Enterprise Document Search; Securing Proprietary Data Ingestion with VPC Service Controls; Customer-Managed Encryption Keys (CMEK) for Sensitive Agent Data Stores
• Consideration 1.2b — Ingesting and processing unstructured multimodal data (e.g., videos, audio, and images) into the agentic workflow: Timestamp-Grounded Video Knowledge Base Ingestion; Customer Call Audio Ingestion with Speaker Diarization; On-the-Fly Image Evaluation via Agent Tools vs. Pre-indexed Datastores; Technical Schematics Ingestion with Layout-Aware Visual Chunking
• Multi-Constraint Production Scenarios, Trade-Off Clinic & Pattern-Trigger Cheat Sheet (1.2)

Chapter 2: Enterprise Coding Agents, Secure Sandboxes & Antigravity Customization

• Subsection 2.1: Using Coding Agents Effectively: MCP Servers, Secure Sandboxes, and Code Optimization (Part II: Using Coding Agents for Application Development (~17% of Exam))
• Domain Overview and Reference Architecture
• Core Foundational Primitives & Default Behaviors (2.1)
• Consideration 2.1a — Configuring coding agents with model context protocol (mcp) servers, custom skills, and access to tools (e.g., antigravity and claude code on google cloud): Secure VPC-Internal MCP Server Integration for Claude Code; MCP Tool Schema Definition for BigQuery Data Inspection; Context Window Management via MCP Tool Response Truncation; Least Privilege IAM Scoping for Coding Agent Execution Tools
• Consideration 2.1b — Using coding agents in secure sandboxes (e.g., google kubernetes engine [gke], cloud workstations, and antigravity): Restricting External Egress for Untrusted Agent-Generated Code in GKE Sandbox; Securing Developer Workspaces with Cloud Workstations and VPC Service Controls; Resource Quotas and Blast Radius Control for Agentic Loops in Antigravity
• Consideration 2.1c — Using coding agents to refactor source code, optimize execution runtimes, and patch application-layer vulnerabilities: Refactoring Blocking I/O to Asynchronous Execution; Remediating SQL Injection via Contextual AST Analysis; Optimizing Algorithmic Time Complexity via Profiling Data; Eliminating N+1 ORM Query Performance Bottlenecks
• Multi-Constraint Production Scenarios, Trade-Off Clinic & Pattern-Trigger Cheat Sheet (2.1)
• Subsection 2.2: Customizing Coding Agents for Enterprise Workflows: Antigravity Skills, Hooks, Plugins, Subagents, and Agents CLI (Part II: Using Coding Agents for Application Development (~17% of Exam))
• Domain Overview and Reference Architecture
• Core Foundational Primitives & Default Behaviors (2.2)
• Consideration 2.2a — Creating skills, plugins, extensions hooks, rules, and subagents using antigravity: Enterprise Security Gate via Pre-Tool Execution Hooks; Enforcing Enterprise Architecture Guidelines via Workspace Rules; Integrating Proprietary Internal APIs as Custom Skills; Multi-Agent Delegation for Security and Code Quality
• Consideration 2.2b — Augmenting antigravity with agents cli to build, scale, govern, and optimize deployed agents: Automated Multi-Environment Build and Packaging; Fleet-Wide Policy and Guardrail Governance; Horizontal Auto-Scaling for Remote Backend Agents
• Multi-Constraint Production Scenarios, Trade-Off Clinic & Pattern-Trigger Cheat Sheet (2.2)

Chapter 3: Custom Agentic Workflows in Code: Models, ADK, Sessions & Memory Bank

• Subsection 3.1: Designing and Building Custom Agentic Workflows in Code: Models, ADK, Sessions, Memory Bank, and CLI Skills (Part III: Developing Custom Agents (~33% of Exam))
• Domain Overview and Reference Architecture
• Core Foundational Primitives & Default Behaviors (3.1)
• Consideration 3.1a — Selecting and configuring the appropriate language model (e.g., large language model [llm] vs. small language model [slm], self-hosted vs. software as a service [saas], and open-source software [oss] vs. proprietary llm) considering cost, security, and agent architecture: Latency and Cost Optimization with SLM Routers; Data Sovereignty and Isolation in Regulated Banking; Predictable Cost Modeling for Continuous High-Throughput Pipelines; Massive Context Analysis for Legal Discovery Agent
• Consideration 3.1b — Building custom agents using open-source libraries (e.g., agent development kit [adk]): Distributed State Persistence in Multi-Turn Agent Workflows; Dynamic OpenAPI Tool Ingestion with Vertex AI Extension Compatibility; Hierarchical Multi-Agent Supervisor Pattern Deployment; Human-in-the-Loop Interrupt and Resume Patterns
• Consideration 3.1c — Configuring sessions and memory (e.g., agent platform memory bank and managed sessions): Short-Term Conversation History vs. Persistent Long-Term Memory; Strict Session TTL and Privacy Compliance; Automatic Fact Distillation for User Profiles
• Consideration 3.1d — Configuring skills using agents cli (e.g., plugins and agent vs. human mode): Interactive Human-in-the-Loop Confirmation for High-Risk Skill Invocation; Configuring Third-Party OpenAPI 3.0 Plugins via Agents CLI; Local CLI Emulation and Mocking of External APIs; Resolving Tool Name Collisions Across Multiple CLI-Registered Plugins
• Multi-Constraint Production Scenarios, Trade-Off Clinic & Pattern-Trigger Cheat Sheet (3.1)

Chapter 4: Enterprise Domain Knowledge: RAG Pipelines, Vector Search, Identity & MCP Tooling

• Subsection 3.2: Integrating Enterprise Domain Knowledge: RAG Pipelines, Vector Search, Agent Identity, and MCP Tool Ecosystems (Part III: Developing Custom Agents (~33% of Exam))
• Domain Overview and Reference Architecture
• Core Foundational Primitives & Default Behaviors (3.2)
• Consideration 3.2a — Designing, configuring, and managing retrieval-augmented generation (rag) pipelines and vector retrieval systems (e.g., embedding models, similarity scoring, and reranking) using appropriate services such as vector databases (e.g., vector search and agent retrieval): Optimizing Chunking Strategies for Tabular Enterprise Policies; Selecting Embedding Task Types for Asymmetric Search; Enhancing Retrieval Precision with Semantic Cross-Encoder Reranking; Mitigating Out-of-Vocabulary Code Lookup Failures with Hybrid Search
• Consideration 3.2b — Configuring agent permissions (e.g., agent identity): End-User Identity Pass-Through vs. Service Account for Document Access; Least Privilege IAM Role Assignment for Grounded Cloud Storage Ingestion; Enterprise Knowledge Grounding with Document-Level ACL Synchronization
• Consideration 3.2c — Using google cloud tools (e.g., agent registry, google cloud mcp servers) to configure prebuilt and custom capabilities (e.g., custom integration layers for managed databases, api integrations, and mcp server that connects agents to third-party saas tools and remote servers): Secure Private Cloud SQL Access via Cloud Run MCP Server; Third-Party SaaS Integration via Remote MCP Server with OAuth Passthrough; Real-Time BigQuery Analytics via Prebuilt Tools vs Custom MCP Server; Centralized Tool Governance with Agent Registry
• Multi-Constraint Production Scenarios, Trade-Off Clinic & Pattern-Trigger Cheat Sheet (3.2)

Chapter 5: Multi-Agent Topologies & Agentic Protocols (MCP vs. A2A)

• Subsection 3.3: Designing multi-agent systems using agentic protocols (Section 3: Designing and Building Production-Grade Custom Agents (33%))
• Domain Overview and Reference Architecture
• Core Foundational Primitives & Default Behaviors (3.3)
• Consideration 3.3a — Orchestrating agents using agentic protocols (e.g., mcp and agent2agent [a2a]): Dynamic Tool Discovery via Remote MCP Server; Context and State Propagation in A2A Communication; Securing MCP Server Access via IAM and OAuth2
• Consideration 3.3b — Selecting and coordinating multiagent handoffs and workflows (e.g., parallel agents, sequential agents, and graph workflow) using google cloud tools (e.g., agent identity, agent registry, agent runtime, and agent policies): Least Privilege Delegation Using Agent Identity in Sequential Handoffs; Parallel Multi-Agent Fan-Out with Vertex AI Agent Runtime Aggregation; Dynamic Capability-Based Routing via Centralized Agent Registry; Cyclic Graph Workflows for Iterative Code Review and Verification
• Multi-Constraint Production Scenarios, Trade-Off Clinic & Pattern-Trigger Cheat Sheet (3.3)

Chapter 6: AgentOps: Evaluation, Autoraters, Deployment, Tracing & Cost Optimization

• Subsection 4.1: Evaluating and testing agents (Section 4: Testing, Deploying, Monitoring, and Scaling Production Agents (22%))
• Domain Overview and Reference Architecture
• Core Foundational Primitives & Default Behaviors (4.1)
• Consideration 4.1a — Creating test sets for agent evaluation (e.g., golden data, prompts, and edge cases): Synthetic Golden Dataset Generation with Human-in-the-Loop Validation; Multi-Turn Context Drift and State Preservation Test Sets; Continuous Test Set Updating from Production Drift Telemetry; Tool-Calling and Parameter Extraction Edge-Case Datasets
• Consideration 4.1b — Creating continuous evaluation pipelines to assess an agent's tool execution based on established success criteria: Automating Production Tool Call Quality Audits; Distinguishing Semantic Tool Argument Errors vs Schema Invalidation; CI/CD Deployment Gating Based on Tool Selection Success Rate
• Consideration 4.1c — Determining the appropriate evaluation framework and tooling (e.g., adk evaluation tooling (evalset), agent platform gen ai evaluation service, and custom autoraters): Local CI/CD Regression Testing with ADK Evalset; Pairwise Model Comparison using Vertex AI Gen AI Evaluation Service; Production Monitoring of Hallucination via Managed Groundedness Metrics; Domain-Specific Regulatory Auditing via Custom Autoraters
• Consideration 4.1d — Evaluating an agentic system against a golden dataset to assess agent response and retrieval quality (e.g., using adk): Retrieval Quality Assessment Using Context Precision and Recall; Agent Tool-Call Trajectory Evaluation Against Golden Sequences; Groundedness and Hallucination Evaluation in RAG-Based Agents; Automated CI/CD Quality Gates for Agent Prompt Iterations
• Multi-Constraint Production Scenarios, Trade-Off Clinic & Pattern-Trigger Cheat Sheet (4.1)
• Subsection 4.2: Deploying, monitoring, and troubleshooting agents (Section 4: Testing, Deploying, Monitoring, and Scaling Production Agents (22%))
• Domain Overview and Reference Architecture
• Core Foundational Primitives & Default Behaviors (4.2)
• Consideration 4.2a — Selecting optimal deployment runtime based on the use case, requirements, and cost (e.g., agent runtime, cloud run, and gke): Serverless Managed Agent Runtime for Python Frameworks; Bursty Tool Calling Microservices with Scale-to-Zero; Low-Latency Agent Workloads with Local Open-Weight Models
• Consideration 4.2b — Troubleshooting agent issues (e.g., drift, tool invocation latency, agent reasoning loops, and system failures): Mitigating Infinite Agent Reasoning Loops via Recursion Limits and Stop Conditions; Troubleshooting High Tool Invocation Latency Using Cloud Trace and Caching; Handling Downstream Tool Failures with Circuit Breakers and Fallback Agents; Detecting and Rectifying Knowledge Base Drift in Grounded Retrieval
• Consideration 4.2c — Monitoring and optimizing agents for performance, reliability, and cost (e.g., identify logic errors, latency bottlenecks, and hallucinations): End-to-End Distributed Tracing for Agentic Latency Bottlenecks; Automated Production Hallucination and Groundedness Monitoring; Cost and Latency Optimization via Semantic Caching; Detecting Tool Call Loops and Logic Failures in Agent Trajectories
• Multi-Constraint Production Scenarios, Trade-Off Clinic & Pattern-Trigger Cheat Sheet (4.2)

Chapter 7: Securing & Governing Agentic Systems: OAuth 2.0, PAB, Model Armor, Guardrails & HITL

• Subsection 5.1: Securing agentic systems (Section 5: Securing and Governing Agentic Systems (15%))
• Domain Overview and Reference Architecture
• Core Foundational Primitives & Default Behaviors (5.1)
• Consideration 5.1a — Implementing authentication and secure tool execution (e.g., agent-to-tool api calls using oauth 2.0): OAuth 2.0 Client Credentials for Internal Service Calls; On-Behalf-Of Delegation via OAuth 2.0 Token Exchange; Cross-Cloud Tool Execution via Workload Identity Federation
• Consideration 5.1b — Configuring principal access boundary (pab) policies using agent identity: Preventing Agent Cloud Storage Data Exfiltration Across Folders; Agent Token Impersonation and Downstream Resource Scoping; Detecting and Auditing PAB Policy Violations in Agentic Tool Execution; Enforcing Multi-Tenant Agent Isolation within a Single Project
• Consideration 5.1c — Configuring agent gateway to monitor traffic and track agents: End-to-End Distributed Tracing Across Multi-Agent Handoffs; Granular Rate Limiting for Autonomous Tool Execution; Mutual TLS and Zero-Trust Agent-to-Tool Authentication; Centralized PII Masking at the Agent Gateway
• Consideration 5.1d — Designing and configuring agentic governance and policy enforcement (e.g., agent registry and model armor): Preventing Indirect Prompt Injections via Model Armor; Governing Multi-Environment Lifecycle via Centralized Agent Registry; Automated PII Redaction in Agent Outputs via Model Armor and Sensitive Data Protection
• Multi-Constraint Production Scenarios, Trade-Off Clinic & Pattern-Trigger Cheat Sheet (5.1)
• Subsection 5.2: Designing safety guardrails and secure data access (Section 5: Securing and Governing Agentic Systems (15%))
• Domain Overview and Reference Architecture
• Core Foundational Primitives & Default Behaviors (5.2)
• Consideration 5.2a — Designing appropriate safety frameworks and guardrails (e.g., agent gateway, model armor, and human-in-the-loop [hitl]): Model Armor Prompt Injection Sanitization; HITL Thresholds for Financial Transactions; Outbound PII and Toxic Output Redaction with Model Armor; Agent Gateway Policy Enforcement for Tool Calling
• Consideration 5.2b — Configuring secure access to data and identity propagation (e.g., agent gateway and agent registry): End-User Identity Propagation via OAuth 2.0 Token Exchange; Cryptographic Verification and Discovery via Agent Registry; Workload Identity Federation for Secure Multi-Cloud Tool Access; Agent Gateway Mediation for In-line DLP and Data Sanitization
• Multi-Constraint Production Scenarios, Trade-Off Clinic & Pattern-Trigger Cheat Sheet (5.2)

Chapter 8: Master Service Disambiguation Matrix & Exam-Day Playbook

• 8.1 The Complete PAA Look-Alike Service Disambiguation Matrix
• 8.2 The 5-Step Exam-Room Decision Protocol
• 8.3 Cross-Cutting Multi-Domain Production Architecture Takeaways (30 Scenarios)
• 8.4 Final Exam-Day Checklist — Top 25 Golden Rules