Was checking out your repo and ran it through Trustabl. The scan found 3 findings, with 2 being medium severity. A couple of things seemed worth mentioning: the Vercel AI tool accepts untyped input, which might be intentional for flexibility, and the Agent configuration appears to be quite opaque. Just thought you'd want to know.
Add Trustabl to your CI — trustabl/trustabl-action:
- name: Trustabl scan
uses: trustabl/trustabl-action@v1
Was checking out your repo and ran it through Trustabl. The scan found 3 findings, with 2 being medium severity. A couple of things seemed worth mentioning: the Vercel AI tool accepts untyped input, which might be intentional for flexibility, and the Agent configuration appears to be quite opaque. Just thought you'd want to know.
Add Trustabl to your CI — trustabl/trustabl-action: