Copilot Instructions for AI Coding Agents

## Project Overview
This repository is for testing the Computer‑Using Agent with Playwright. It is minimal and primarily serves as a sandbox for agent-driven automation and integration testing.

## Architecture & Key Components
- The codebase is intentionally sparse. The main focus is on agent workflows and Playwright-based automation.
- There are no complex service boundaries or data flows. The structure is flat, with most logic expected to be added by agents or during test runs.

## Developer Workflows
- **Testing:** Playwright is the primary tool for browser automation and testing. Use Playwright commands to run, debug, and author tests.
- **Setup:** If Playwright is not installed, run `npm install playwright` or use the Playwright CLI for setup.
- **Test Execution:** Typical test commands:
  - `npx playwright test` — runs all Playwright tests
  - `npx playwright codegen` — launches Playwright code generator
- **Debugging:** Use Playwright's built-in debugging tools (e.g., `PWDEBUG=1 npx playwright test`).

## Project-Specific Patterns
- There are no custom conventions or patterns yet. Follow standard Playwright and Node.js practices unless otherwise specified in future commits.
- If adding new files, prefer clear, descriptive names and keep the structure flat unless complexity demands otherwise.

## Integration Points
- Playwright is the only external dependency currently referenced. Integrate additional tools only if required for agent testing.
- No cross-component communication patterns are present.

## Key Files
- `README.md`: Project purpose and minimal setup notes.
- `.github/copilot-instructions.md`: This file — update as the project evolves.

## Example: Adding a Playwright Test
```
bash
npx playwright test
```

## Guidance for AI Agents
- Assume minimal context; most logic will be agent-generated.
- Document any new conventions or workflows in this file as the project grows.
- If you add new dependencies or tools, update this file with usage instructions.