Apppublic

Sukrit789/shipping-support-agent

sourceHugging Faceupdated 3mo agoView on Hugging Face

0likes

App README

shipping-agent

Agent generated with agents-cli version 0.5.0

🤖 What This Agent Does

This project implements an automated Customer Support Representative agent for a shipping company using the Google Agent Development Kit (ADK) 2.0.

1.Intelligent Query Classification: When you ask a question, a specialized classification agent analyzes your query to check if it's related to shipping (such as tracking, rates, delivery times, or returns).
2.Smart Routing:
3.Shipping Queries are routed to a specialized Shipping FAQ agent. Responses concerning shipping rates are enthusiastic, filled with fun emojis, and highlight the FREE shipping threshold on orders over $50.00! 🚚🚀🎉
4.Unrelated Queries (e.g., general knowledge questions) are routed to a decline node that politely directs the user back to shipping-related topics.

🤝 Multi-Agent Architecture

This project is built as a multi-agent system using a directed workflow graph. Instead of a single LLM trying to do everything, tasks are split between specialized agents and programmatic logic:

mermaid

graph TD
    START([START]) --> ProcessStart[1. Process Start Node]
    ProcessStart --> Classifier[2. Classifier Agent]
    Classifier --> Router[3. Router Function]
    Router -- "shipping" --> FAQAgent[4a. Shipping FAQ Agent]
    Router -- "unrelated" --> DeclineNode[4b. Decline Node]
    FAQAgent --> END([END])
    DeclineNode --> END([END])

—`classifier` Agent: A dedicated LlmAgent tasked solely with classifying user intents against a structured Pydantic schema.
—`shipping_faq_agent` Agent: A dedicated domain-specific LlmAgent trained on company policies to provide policy answers in a friendly, enthusiastic style.
—Workflow Orchestration: Structural Python nodes (process_start, router_node, and decline_node) control the routing and execution flow based on output states.

Project Structure

customer-support-agent/
├── app/         # Core agent code
│   ├── agent.py               # Main agent logic
│   └── app_utils/             # App utilities and helpers
├── tests/                     # Unit, integration, and load tests
├── GEMINI.md                  # AI-assisted development guide
└── pyproject.toml             # Project dependencies

💡 Tip: Use Gemini CLI for AI-assisted development - project context is pre-configured in GEMINI.md.

Requirements

Before you begin, ensure you have:

—uv: Python package manager (used for all dependency management in this project) - Install (add packages with uv add <package>)
—agents-cli: Agents CLI - Install with uv tool install google-agents-cli
—Google Cloud SDK: For GCP services - Install

Quick Start

Install agents-cli and its skills if not already installed:

bash

uvx google-agents-cli setup

Install required packages:

bash

agents-cli install

Test the agent with a local web server:

bash

agents-cli playground

You can also use features from the ADK CLI with uv run adk.

Commands

Command	Description
`agents-cli install`	Install dependencies using uv
`agents-cli playground`	Launch local development environment
`agents-cli lint`	Run code quality checks
`agents-cli eval`	Evaluate agent behavior (generate, grade, analyze, and more — see `agents-cli eval --help`)
`uv run pytest tests/unit tests/integration`	Run unit and integration tests

🛠️ Project Management

Command	What It Does
`agents-cli scaffold enhance`	Add CI/CD pipelines and Terraform infrastructure
`agents-cli infra cicd`	One-command setup of entire CI/CD pipeline + infrastructure
`agents-cli scaffold upgrade`	Auto-upgrade to latest version while preserving customizations

Development

Edit your agent logic in app/agent.py and test with agents-cli playground - it auto-reloads on save.

Deployment

bash

gcloud config set project <your-project-id>
agents-cli deploy

To add CI/CD and Terraform, run agents-cli scaffold enhance. To set up your production infrastructure, run agents-cli infra cicd.

Observability

Built-in telemetry exports to Cloud Trace, BigQuery, and Cloud Logging.