Feat: Shortest

[m2rads]

AI-Powered E2E Testing Framework

Addresses: #69

This PR introduces a complete end-to-end testing framework powered by Claude AI. The framework enables writing tests in natural language and executes them through AI-driven browser automation.

Key Features

• Natural language test definitions using a fluent API
• AI-powered test execution using Claude computer use
• Automated browser interaction with precise coordinate handling
• Detailed test reporting with pass/fail status

Technical Highlights

• Direct integration with Anthropic's Claude API
• Playwright-based browser automation
• Coordinate scaling system for reliable element targeting
• Visual mouse tracking for debugging
• Modular architecture with clear separation of concerns

[vercel[bot]]

@m2rads is attempting to deploy a commit to the Antiwork Team on Vercel.

A member of the Team first needs to authorize it.

[m2rads]

Hi @slavingia! Please do not merge this PR just yet. I noticed an issue with Vercel deployment. Let me resolve that first.

Please clone this branch locally if you'd like to test the lib.

[m2rads]

Documentation is in docs/README.md

[m2rads]

Hi @slavingia looking at my Vercel looks like the deployment was done successfully. Can you check on your side to see if there are any issues?

[m2rads]

Hi @slavingia, addressed your comments in this commit. Looking forward to further code review.

[m2rads]

Should I add unit tests for this package? There are some test scripts available but do not cover all areas.

[slavingia]

Merged but if you wish to add unit tests as a new PR you're welcome to do so. I like the idea of trying to rely 100% on shortest tests.