cancel stop loading window

[LawyZheng]

🚀	This description was created by Ellipsis for commit 4c22bdb

Summary:

Replaced STOPLOADING scrape type with an additional NORMAL scrape type in the scraping retry logic.

Key points:

• Modified SCRAPE_TYPE_ORDER in skyvern/constants.py to replace ScrapeType.STOPLOADING with ScrapeType.NORMAL.
• Updated _build_and_record_step_prompt in skyvern/forge/agent.py to retry normal scraping twice before reloading the page.
• Changed error logging message in _build_and_record_step_prompt to reflect the new retry logic.

---

Generated with ❤️ by ellipsis.dev

[ellipsis-dev]

❌ Changes requested. Reviewed everything up to 4c22bdb in 52 seconds

More details

• Looked at 51 lines of code in 2 files
• Skipped 0 files when reviewing.
• Skipped posting 0 drafted comments based on config settings.

Workflow ID: wflow_lotJw86ZEy8vXwnO

---

Want Ellipsis to fix these issues? Tag @ellipsis-dev in a comment. You can customize Ellipsis with 👍 / 👎 feedback, review rules, user-specific overrides, quiet mode, and more.

[ellipsis-dev]

The change to have two ScrapeType.NORMAL entries in SCRAPE_TYPE_ORDER might not be effective for different scraping strategies. Consider reintroducing ScrapeType.STOPLOADING or another distinct strategy to maintain diverse scraping attempts.

Suggested change

	SCRAPE_TYPE_ORDER = [ScrapeType.NORMAL, ScrapeType.NORMAL, ScrapeType.RELOAD]
	SCRAPE_TYPE_ORDER = [ScrapeType.NORMAL, ScrapeType.STOPLOADING, ScrapeType.RELOAD]

[wintonzheng]

if we're removing STOPLOADING here, can we also remove https://github.com/Skyvern-AI/skyvern-cloud/blob/82c2f11926c12a57ea592ecd8c2b33fdededb73d/skyvern/forge/agent.py#L846
?

[LawyZheng]

Can we keep it? I suspect it stuck the task due to the no timeout for page.evaluate(). Maybe we will be back to use it again when we find a way to time out page.evaluate()