Retail Agent Customizer

Name: retail-agent-customizer
Rating: 76
Author: lavinigam-gcp

Common Customizations

Goal	What to Modify
Different business type	IntakeAgent + MarketResearchAgent prompts
New industry vertical	All agent prompts + StrategyAdvisor criteria
Additional output format	Add new agent to ParallelAgent
Different analysis criteria	GapAnalysisAgent + report schema
Change AI model	`app/config.py`

Quick Customizations

Change Business Vertical

Edit IntakeAgent to recognize new business types:

python

# app/sub_agents/intake_agent/agent.py
INSTRUCTION = """Extract:
- target_location: The location/city/neighborhood
- business_type: Type of business (e.g., restaurant, gym,
  salon, clinic, coworking space, YOUR_NEW_TYPE)
"""

Modify Research Focus

Edit MarketResearchAgent for industry-specific research:

python

# app/sub_agents/market_research/agent.py
INSTRUCTION = """Research the following for {target_location}:
- Demographics and foot traffic
- [ADD YOUR CRITERIA HERE]
- Industry-specific trends for {business_type}
"""

Change Output Schema

Edit the Pydantic schema for different report structure:

python

# app/schemas/report_schema.py
class LocationIntelligenceReport(BaseModel):
    location_score: float
    market_opportunity: str
    # Add your custom fields
    your_custom_field: str

Add New Output Format

Create new artifact agent and add to ParallelAgent:

python

# app/agent.py
artifact_generation_pipeline = ParallelAgent(
    name="ArtifactGenerationPipeline",
    sub_agents=[
        report_generator_agent,
        infographic_generator_agent,
        audio_overview_agent,
        your_new_output_agent,  # Add here
    ],
)

Key Files to Modify

Customization	Primary Files
Business types	`app/sub_agents/intake_agent/agent.py`
Research focus	`app/sub_agents/market_research/agent.py`
Competitor criteria	`app/sub_agents/competitor_mapping/agent.py`
Analysis logic	`app/sub_agents/gap_analysis/agent.py`
Recommendations	`app/sub_agents/strategy_advisor/agent.py`
Report structure	`app/schemas/report_schema.py`
Model selection	`app/config.py`
Output artifacts	`app/sub_agents/artifact_generation/agent.py`

[See references/customization-guide.md for detailed examples]