Craigslist-tester-agent

Lightweight automation utilities and example agents for driving a browser-based QA workflow (used with the browser-use automation agent).

This repository contains small scripts and a sample feature describing a Craigslist QA task (navigate to Miami Craigslist, find the 'boats' listing, and verify filtering behavior). It's intended as a compact playground for browser-driven acceptance testing and experiment with the browser-use agent.

Quick overview

• Project: craigslist-tester-agent
• Purpose: automation scripts and artifacts that demonstrate driving a browser to execute acceptance-style scenarios against websites (example: Craigslist boats-for-sale).
• Primary files:
  • boats-for-sale.feature.md — feature scenarios to validate on miami.craigslist.org
  • agent.py — example runner/agents (small scripts)

Requirements

• Python 3.13+ (project pyproject.toml indicates >=3.13)
• uv package manager (required for environment setup and running scripts)

Environment variables

You may need to provide API keys or other configuration via environment variables. The recommended approach is to create a .env file in the project root and set any required variables there. For example:

# .env
BROWSER_USE_API_KEY=your_api_key_here

Alternatively, you can set environment variables directly in your shell before running the agent scripts. Make sure any required secrets or configuration are available to the agent at runtime.

Getting started

1. Clone the repo and open it in VS Code.
2. Use uv to set up your Python environment and sync dependencies:

uv venv
source .venv/bin/activate
uv sync

3. Review the boats-for-sale.feature.md to understand the acceptance scenarios to run. The repository contains small example scripts (main.py, agent.py) that show how an automation agent could be used to run the scenarios.

Running the example

This repository is a demonstration; there is no single turnkey CLI included. Typical steps to exercise the examples:

1. Start the browser-use automation server (per your local setup).
2. Run the agent script to start the scripted agent logic that navigates to craigslist and runs the scenarios. For example:

uv run agent.py

Expect the agent to open a browser session, navigate to https://miami.craigslist.org, and perform the checks described in boats-for-sale.feature.md. The exact output depends on your automation backend configuration.

Files of interest

• boats-for-sale.feature.md — feature file describing the QA scenarios.
• README.md — this file.

How to contribute

This repo is a personal/sandbox project. Small improvements are welcome: add tests, improve the feature file, or wire a more robust runner. When in doubt, open an issue listing what you'd like to change.

---