botpress

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

• Third-party content exposure detected (medium risk: 0.30). The workflow involves interacting with Botpress via the Membrane CLI, which can read messages and user content from chat conversations and data tables.

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

• Potentially malicious external URL detected (high risk: 0.70). The skill instructs installation and execution of the Membrane CLI from npm (@membranehq/cli), which is an unversioned third-party package identifier without a direct repository locator, constituting an unverifiable runtime external dependency.