adaptyv

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.30). The workflow reads API responses, experiment results, and lists from the Adaptyv Bio Foundry API, which can contain user/organization-submitted protein sequences, experiment names, and metadata from external sources.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (medium risk: 0.30). The skill instructs installing a Python package from a GitHub repository via uv pip install git+https://github.com/adaptyvbio/adaptyv-sdk.git..., which is an operational dependency fetched from a remote source.

MEDIUM W009: Direct money access capability detected (payment gateways, crypto, banking).

• Direct money access detected (high risk: 0.70). The skill operates the Adaptyv Bio Foundry API and Python SDK to design protein characterization experiments, estimate costs, submit sequences, monitor laboratory progress, and retrieve results. It supports estimating costs, confirming quotes, and creating invoices for laboratory experiments (direct financial operations/purchases), as well as automated billing options like auto_accept_quote. Authentication is handled via bearer tokens (ADAPTYV_API_KEY) checked against the API, but since these are direct purchases/service authorizations without independent transaction-time user approval or second-factor/independent payment verification enforced by the execution layer beyond API key presentation, it falls under direct operations without established qualifying control.