vertical-onboarding

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

• Third-party content exposure detected (high risk: 0.70). The required workflow is “import-first” via migration-import-engineer’s import contract, which at runtime necessarily ingests the user’s chosen incumbent data (outsider-authored content from the incumbent/service being imported) into the LLM context for preview/verification, creating an indirect prompt-injection risk.