Environment: Quick Web App (Flows feature) When it broke: Noticed ~July 1, 2026 (was working as recently as mid-June 2026) Impact: Cannot create new flows — completely blocks flow development
Issue: When creating a new flow and pasting a main prompt that previously generated the correct steps automatically, the web app now only outputs a generic/default flow template. It does not parse the prompt to create the defined steps.
What I’ve tried:
Pasting the exact same prompt that successfully generated 12 correct steps previously (7 file upload steps, 1 text input step, 4 code execution steps) — now produces generic template
Restructuring the prompt with steps listed explicitly at the top in various formats — same result
Simplifying formatting (removing markdown tables, headers) — same result
Creating a brand new flow from scratch — same result
Expected behavior: The flow auto-generator reads the main prompt, identifies the step definitions, and creates the corresponding steps with the correct types (file upload, text input, code execution). This worked correctly for weeks prior to this issue.
Actual behavior: Regardless of prompt content or formatting, the app generates a default/generic flow template that does not reflect the steps defined in the prompt.
Questions:
Was there a recent update to the flow step generation logic?
Is there a new required format for step definitions in the main prompt?
Is there an alternative method to manually create steps that I’m missing?
This is blocking an active hackathon project (guard hour reconciliation tool for DC Security). Any guidance appreciated.
This sounds like a service-side regression rather than anything you’re doing wrong. If the exact same prompt was generating the correct steps up through mid-June and now it’s only producing a generic template regardless of what you pass in, that points to a change on the backend rather than a prompt formatting issue.
If you can, I would recommend creating a support ticket and the support team will be able to further investigate into this issue. Please refer to this resource on how to create a support ticket (Case management - AWS Support).
However, if this hackathon is sponsored/organized by AWS, then I would try reaching out to someone from there about this and seeing if they can get in contact with anyone.
Thank you very much for the reply. I’ll go ahead and talk to the hackathon team regarding this issue to see if they can contact someone. If not I’ll proceed with the ticket submission.
Just checking back in since we haven’t heard from you in a bit. I wanted to see if the guidance shared earlier helped resolve your question, or if you found a solution in the meantime. Moreover, if you submitted a support ticket and it uncovered any findings, please feel free to share them!
If you still have any additional questions related to your initial post, feel free to share them. Otherwise, any update you’re able to provide within the next 3 business days would be helpful for the community.
It did not help. I’m an internal AWS employee, but I do not work with any AWS platform accounts. The ticket link that was shared was for AWS platform support. So not sure how to go about this or if there is an internal track we can use to get help?
Unfortunately, the community is a public facing forum and doesn’t have visibility into internal escalation channels. What I’d suggest is reaching out to your team’s AWS account owner or IT admin. If you do find a resolution, sharing it here would be helpful for the community!