Request: Browser-compatible bundle for Java/GraalVM integration

### Title
Request: Browser-compatible bundle for Java/GraalVM integration

### Description
I'm working on a Java wrapper for the AsyncAPI parser using GraalVM's JavaScript engine. The current @asyncapi/parser package (v3.6.0) cannot be bundled for use in GraalVM because it depends on Node.js-specific APIs (fetch, Headers, etc.) that aren't available in GraalVM's JavaScript environment.

### Current Situation
- The old bundle-es5.js only supports AsyncAPI 2.4.0
- The new @asyncapi/parser v3.x requires fetch API and browser globals
- When bundled with webpack, it fails with: `TypeError: Cannot set property 'Headers' of undefined`

### Request
Could you provide or help create a browser-compatible bundle that:
1. Supports both AsyncAPI v2.x and v3.x
2. Works in environments without fetch API (or includes polyfills)
3. Can run in GraalVM's JavaScript engine
4. Doesn't depend on Node.js-specific modules

### Use Case
Java applications that need to parse AsyncAPI documents without spawning Node.js processes or using native V8 bindings (which don't support macOS ARM).

### Alternative Solutions Considered
- Using Playwright/Puppeteer (too heavy)
- Using J2V8 (no macOS ARM support)
- Creating custom bundle with webpack (fails due to fetch dependency)

### Environment
- GraalVM JS engine via javax.script
- Target: Java 11+
- Platform: macOS ARM, Linux, Windows

Would appreciate guidance on:
1. How to properly bundle the parser for non-browser/non-Node environments
2. If there's an existing standalone bundle
3. If this use case is supported

Start by inspecting the current @asyncapi/parser v3.6.0 package and the existing bundle-es5.js, then reproduce the webpack failure involving fetch and Headers. Determine how the parser’s Node.js-specific dependencies enter the bundle and whether both AsyncAPI v2.x and v3.x can share a supported output. Done means a documented bundle runs in GraalVM’s JavaScript engine without Node.js APIs and covers both requested versions.

Written by the indexing model from the issue text.