retail-agent-learner
Learn ADK concepts and understand the retail agent pipeline. Use when asking how the agent works, what ADK is, wanting to understand the architecture, state flow, or individual agent purposes.
When & Why to Use This Skill
This Claude skill serves as a comprehensive educational resource for mastering the Agent Development Kit (ADK) and multi-agent system architecture. It provides a deep dive into the 'retail-agent-learner' pipeline, illustrating how to structure complex AI workflows, manage state flow between specialized agents, and implement both sequential and parallel execution patterns. It is an essential guide for developers looking to design robust, scalable agentic solutions using industry-standard patterns.
Use Cases
- Architecting Multi-Agent Pipelines: Learn how to decompose a complex business problem (like retail site selection) into a structured 8-agent workflow involving specialized roles like MarketResearch and GapAnalysis.
- Mastering State Management: Understand how to pass data and context between different agents using session state and ToolContext to ensure continuity in long-running tasks.
- Implementing Execution Patterns: Study the practical differences and implementation details of SequentialAgent vs. ParallelAgent for optimizing performance and concurrency.
- Developer Onboarding for ADK: Utilize the 9-part tutorial series to transition from basic agent setup to production-level deployment, including testing and artifact generation.
| name | retail-agent-learner |
|---|---|
| description | Learn ADK concepts and understand the retail agent pipeline. Use when asking how the agent works, what ADK is, wanting to understand the architecture, state flow, or individual agent purposes. |
Retail Agent Learner
Pipeline Overview
This is an 8-agent pipeline for retail site selection analysis:
User Query
↓
IntakeAgent ────────────→ Extract location + business type
↓
MarketResearchAgent ────→ Google Search for demographics
↓
CompetitorMappingAgent ─→ Google Maps Places API
↓
GapAnalysisAgent ───────→ Python code execution (pandas)
↓
StrategyAdvisorAgent ───→ Extended thinking synthesis
↓
ParallelAgent ──────────→ Concurrent artifact generation
├── ReportGenerator ─→ HTML report
├── InfographicAgent → Image generation
└── AudioOverview ───→ TTS podcast audio
What Each Agent Does
| Agent | Purpose | Key Tool/Feature |
|---|---|---|
| IntakeAgent | Parse user request | AgentTool pattern |
| MarketResearchAgent | Find market data | google_search built-in |
| CompetitorMappingAgent | Map competitors | Google Maps API |
| GapAnalysisAgent | Calculate viability | Code execution (pandas) |
| StrategyAdvisorAgent | Synthesize recommendations | Extended thinking |
| ReportGenerator | Create HTML report | Artifact generation |
| InfographicAgent | Generate visual | Image generation |
| AudioOverview | Create podcast audio | TTS multi-speaker |
Key ADK Concepts
Agents
- LlmAgent: Single LLM call with tools and instructions
- SequentialAgent: Runs sub-agents one after another
- ParallelAgent: Runs sub-agents concurrently
State Flow
Data passes between agents via session state:
IntakeAgent → state["target_location"], state["business_type"]
MarketResearchAgent → state["market_research_findings"]
CompetitorMappingAgent → state["competitor_analysis"]
GapAnalysisAgent → state["gap_analysis"]
StrategyAdvisorAgent → state["strategic_report"]
Tools
Functions that agents can call to perform actions. Access state via ToolContext.
Callbacks
Lifecycle hooks: before_agent_callback and after_agent_callback.
Learning Path
Start with the 9-part tutorial series:
- Part 1: Setup + First Agent
- Part 2: IntakeAgent
- Part 3: MarketResearchAgent
- Part 4: CompetitorMappingAgent
- Part 5: GapAnalysisAgent (code execution)
- Part 6: StrategyAdvisorAgent (extended thinking)
- Part 7: ArtifactGeneration (parallel outputs)
- Part 8: Testing
- Part 9: Production Deployment
[See references/architecture.md for detailed data flow] [See references/adk-concepts.md for ADK deep dive]