On our team we are making a bash script that downloads the data from PB using curl. We are currently using the Arc-Token cookie which saves the Okta session and expires every 24 hours. Then it would be very useful if the PB data is available on some endpoint so that an access token created in the Developer Center section can be used.
Martin Rouede
Hi John,
Can you elaborate more about what context is missing?
pb-data is essentially mongodump output compressed (tar.gz). Which is MongoDB's binary jsonl format: bson.
If you uncompress that package, you'll notice bson files. If you have mongodb installed in your system, it's CLI tools includes a utility to convert bson files to jsonl files which then can be easily scripted or analyzed (although some files may be really big, like 1gb+ for some customers).
We use this method to provide this recipe and scripts to analyze pb-data contents: https://dev.arcxp.com/pagebuilder-engine/how-to-guides/optimization-and-deployment/how-to-check-featurecontent-source-usage-using-pb-data-analysis-scripts/
These scripts were created before Claude code or current frontier models being this capable. I'd easily see claude-code, codex, or your favorite agentic ide, being able to manipulate to surface the context you need.
If there are still missing context you need we'd like to hear that, also a good candidate for a separate feature request.
Having the configuration context provided in that download, which is not available in the repo alone, can be very useful not only in diagnosing rendering problems on sites with many different page and template configurations but it can also provide valuable context to AI coding tools about which templates and pages use which components and in which combination and order. Without that, it can be difficult for AI systems to understand the full scope and context of code changs. At a bare minimum, getting the component, template, and page information in a machine-readable format would be a big plus.
+1 to Eckie's comment. Martin Rouede's comment from way back about having the use in localhost would really be helpful still as well - This would be a solid win. Would love to see this move forward.
This feature would be most useful especially in the AI age now. The pb data resulting bson/json files provide valuable context combined with the Fusion repo for higher quality code, insights, and analysis.
Hi Martin, I agree that the PB Data Sync API provides related but different functionality.
The title and the description of this feature request is distinctly different, so I'm marking it back to Future consideration as we are currently reviewing pb-data and PB Editor programmatic access related ideas.
Hi Jenae, sounds great, but this API doesn't work on localhost. Our use case is to perform these 2 steps automatically via an in-house script instead of clicking on a UI and navigating the file system. (See the attached file).
Hi Martin,
We now offer the ability to copy your PageBuilder data via API.
Please refer to the ALC Documentation on Syncing PageBuilder data across environments and PageBuilder Editor Data Sync API to get started.
As always, feel free to reach out to your TAM if you have questions.