formal-methods-drift-guard

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

• Third-party content exposure detected (high risk: 0.75). The required workflow ingests outsider-authored free text via the “Trace/log replay” and “Text/code diff” evidence paths (e.g., CI logs, incident traces, and diffed docs/code/model files) that are not authored by the operating user, and then translates that content into ledger/domain wording for the LLM context.