## Overview

High Dimensional Research announced Nolita 2.0, describing it as "a natural language, drop-in Page API." The tool is accessible via GitHub or through the command `npx nolita@latest`.

## Key Developments

### API Redesign

Following a presentation at Betaworks' Agent Camp Demo Day, the team restructured Nolita around `Page` instances. This architecture enables developers to work with "the same browser session -- not just one-off objective sessions."

### Core Takeaways

The team identified three important insights:

- "objective-oriented, agentic navigation isn't always the ideal for quick tasks"
- Developers using Python deserve support despite TypeScript codebase
- Drop-in replacement functionality matters for usability

### API Methods

**`Page.do`**: Executes single, action-oriented commands

**`Page.get`**: Accomplishes objectives with typed response expectations

Example code demonstrates navigating to a URL, clicking elements, and extracting structured data using Zod schemas.

## Documentation Expansion

Enhanced inline documentation now appears across exported classes, with a dedicated [documentation site](https://docs.nolita.ai/) showing definitions, parameters, and usage examples.

## Product Naming

The `@hdr/browser` package was officially named Nolita, inspired by Manhattan's Nolita neighborhood.

## Future Plans

- Remote Chrome instance abstraction for `Page` objects
- Memory Index soft onboarding and expanded documentation
- Python SDK development