custom logic for status codes

You can check the status code in the response object from the PuppeteerCrawlingContext and do some logic based on that: https://crawlee.dev/api/puppeteer-crawler/interface/PuppeteerCrawlingContext#response

For example if the status code is 403 it gets retried before I have the chance to check this response object

In the session pool options, you can specify what status codes should be treated as blocked requests. I think 403 is there by default, so passing an empty array will allow you full control here. Look this up - https://crawlee.dev/api/core/interface/SessionPoolOptions#blockedStatusCodes