Choose your source
Published URL
Your coding agent opens the public site in a browser and rebuilds what it sees. Public pages expose less than source
files do.
Source files
Give your coding agent a repository, export, structured content, or media files to use as migration evidence.
Set up the target
Run one of:fimo-migration skill, reads the source itself, and imports the result with the normal fimo commands. See the CLI reference for every setup flag.
Review the result
Before you treat a migration as complete, compare the rebuilt site with the source. Check every route, content collection, asset, form, locale, redirect, and metadata field that matters to your site. Record anything missing or unsupported instead of assuming the migration preserved it. The migration skill’s QA gate checks the rendered site, not the repository. Your coding agent lists the routes and the brand, legal, and navigation rules the site must keep, then proves each rule on rendered pages. Once copy lives in Fimo, a search of the source files cannot show what visitors see, and a check that observes no pages or no links fails instead of passing. The migration report includes that result; treat a missing or failed guardrails section as an incomplete migration. Saved browser captures must include a successful page response and its final URL. Missing response evidence or a reported navigation error fails the check. Internal navigation checks include relative links and links on the same origin. Your coding agent still needs to review the pages in a browser, including error pages served with a successful status, links to specific page sections, and external destinations. Preserve needed redirects in Fimo’s hosting config. Migration does not publish the result. Follow the normal deploy and publish workflow only after you have reviewed the site.What’s next
Source limitations
Decide which source material to provide and what a public page cannot recover.
Migrate from source
Use an existing repository or export as evidence for your coding agent.