Reframe FAQ 'Unsupported Websites' as bot-detection education#459
Conversation
The flat "Unsupported Websites" list overstated reality: Kernel browsers navigate these sites fine. Friction shows up on login flows, high-volume scraping, and aggressive-vendor sites, not on page loads. Replace the list with guidance on where resistance actually occurs and which mitigations to use. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
|
Preview deployment for your docs. Learn more about Mintlify Previews.
💡 Tip: Enable Workflows to automatically generate PRs for you. |
![cursor[bot]](https://avatars.mcszy.xyz/in/1210556?s=60&v=4){kind=small}
There was a problem hiding this comment.
Risk assessment: Very Low
This changes one FAQ MDX file (11 additions, 8 deletions) and only updates user-facing documentation copy plus links to existing documentation pages. It doesn't modify runtime logic, configuration, infrastructure, security boundaries, or model prompts. The changed surface is narrow, and the broken-links and Mintlify checks passed.
Sent by Cursor Automation: Assign PR reviewers
Drop login/account-creation from the "where you'll hit resistance" list so it doesn't frame login as a limitation. Managed Auth handles that, so point to it positively instead. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
There was a problem hiding this comment.
Risk assessment: Very Low
The updated diff still changes only browsers/faq.mdx: it removes one FAQ bullet and rephrases the Managed Auth guidance. The PR modifies documentation copy and links only, with no runtime logic, configuration, infrastructure, security boundary, or model-prompt changes. The latest update doesn't increase the narrow user-facing blast radius.
The PR already has an approval, so I haven't submitted another approval.
Sent by Cursor Automation: Assign PR reviewers
Remove any claim about whether specific sites can or cannot be automated, and stop naming individual sites. Explain that detection is site- and configuration-specific and what drives it. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
masnwilliams
changed the title
There was a problem hiding this comment.
Risk assessment: Very Low
The current diff changes only browsers/faq.mdx (12 additions, 8 deletions), replacing FAQ copy and adding links to existing documentation pages. It doesn't modify runtime code, configuration, infrastructure, security boundaries, or model prompts. The update since the approved commit remains limited to documentation wording, so the risk hasn't increased.
I also verified that all six newly referenced local documentation routes exist and that the diff passes git diff --check. The PR already has an approval, so I haven't submitted another approval.
Sent by Cursor Automation: Assign PR reviewers
|
Preview deployment for your docs. Learn more about Mintlify Previews.
💡 Tip: Enable Workflows to automatically generate PRs for you. |
Brings in the latest content edits from main — the FAQ 'Unsupported Websites' → bot-detection reframing (#459) and the per-browser-type Stripe pricing updates. Rename detection carried them onto the migrated content/docs paths; no code overlap, build + lint green. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
1 participant
Summary
The FAQ's Unsupported Websites section listed specific sites (LinkedIn, Facebook, Instagram, X, Amazon, Reddit) as unsupported. That was inaccurate and had caused customer-facing confusion.
This rewrites the section as education rather than a supported/unsupported verdict. It makes no claim about whether any specific site can or cannot be automated — instead it explains that bot-detection behavior is site- and configuration-specific, what drives friction, and how to reduce it.
What changed
Replaces the flat list with a "Bot detection varies by site" section that:
Notes
🤖 Generated with Claude Code
Note
Low Risk
Documentation-only copy and link updates with no runtime or API behavior changes.
Overview
The browsers FAQ no longer names LinkedIn, Facebook, Instagram, X, Amazon, and Reddit as unsupported. The Unsupported Websites section becomes Bot detection varies by site, explaining that friction depends on approach and scale rather than a binary supported/unsupported list.
It adds practical drivers of detection—high-volume scraping from shared exit IPs and aggressive vendors (Cloudflare, DataDome, PerimeterX, Imperva, Akamai)—with pointers to proxies, Profiles, stealth mode, computer controls, and Managed Auth for logged-in flows. An
<Info>callout directs readers to test targets manually and use the bot detection guide.Reviewed by Cursor Bugbot for commit ba58603. Bugbot is set up for automated code reviews on this repo. Configure here.