Migrating an existing website
Complete guide to importing an existing site into bldrAgent, including host settings, login credentials, best setup patterns, and common migration failures to avoid.
Migrating an existing website into bldrAgent
Use website import when you already have a live site and want to rebuild or modernize it in bldrAgent with as much structure preserved as possible.
This flow can import pages, copy media, and preserve content patterns so you can iterate faster with AI after import.
Before you start
Prepare these items first:
- production site URL,
- backend login credentials when required,
- list of priority pages (home, pricing, contact, etc.),
- known sections to skip (private account areas, checkout, legacy landing pages),
- a short QA checklist for post-import validation.
If your current platform supports app passwords or API keys, prefer those over your main account password.
Import settings explained
When creating your import, you can tune these settings:
| Setting | What it controls | Recommended starting value |
|---|---|---|
| Base URL | Site origin to import from | Your canonical primary domain |
| Max pages | Total pages pulled in one run | 12 to 20 |
| Crawl depth | How far links are followed | 1 (increase to 2 only if needed) |
| Max assets | Number of images/files imported | 220 |
| Include paths | Only import specific sections | Add your highest-value paths |
| Exclude paths | Skip sections that should not migrate | Add auth, cart, checkout, or admin paths |
| Website host type | Import strategy for known platforms | Select exact host when known |
| Backend username/password | Credentials for protected APIs/content | Use least-privileged credentials |
Best setup by host type
Choose the closest website host type in the importer:
- WordPress: best when you can provide WordPress username + application password. This allows direct API reads for posts/pages and usually improves content quality.
- Shopify: choose when storefront data structure is important and you need collection/product context captured.
- Webflow: choose when pages rely on CMS structures and layout conventions from Webflow.
- Wix / Squarespace / Other: crawler-first import. Works well for public pages, but dynamic sections may need manual cleanup.
If you are unsure, use Auto first. If output quality is low, rerun with an explicit host type.
Credentials: what to enter
For sites that support authenticated import:
- Enter the backend username
- Enter the backend password or app password
- Optionally provide API key when your platform requires it
- Keep credentials scoped to read-only access where possible
Do not use owner-level credentials if a restricted migration account is available.
Recommended migration workflow
- Run a focused import of your top 10 to 20 pages
- Verify layout + content quality on priority pages first
- Fix structure and branding with AI prompts
- Expand import scope in a second pass only if needed
- Finalize SEO metadata, links, and forms before publishing
This phased approach reduces risk and avoids large low-quality imports that are expensive to clean up.
What can go wrong (and how to recover)
1) Site blocks scraping
Some websites block automated crawling with bot protection, WAF rules, robots settings, or geo restrictions.
What to do:
- try authenticated import with credentials,
- reduce crawl depth and target specific include paths,
- whitelist the importer where your host supports it,
- migrate priority pages manually if blocking cannot be removed.
2) Private or dynamic sections fail
Account dashboards, carts, and script-heavy widgets may not import cleanly.
What to do:
- exclude those paths during import,
- rebuild dynamic flows directly in bldrAgent,
- use AI prompts to recreate behavior from requirements.
3) Branding or media mismatches
Imported CSS and media can look different than expected.
What to do:
- run a post-import branding cleanup pass,
- replace critical assets in media manager,
- standardize typography/colors in global settings.
4) Incomplete page coverage
Large sites may exceed the first-run page and asset limits.
What to do:
- import in batches by section,
- use include paths for controlled runs,
- track migrated URLs in a checklist.
Post-import validation checklist
- navigation links resolve correctly,
- forms submit and route as expected,
- images are loading from imported media,
- metadata/title structure is updated,
- mobile layout is verified on key pages,
- blocked or failed pages are documented for rebuild.