portfolio-canvas

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

• Third-party content exposure detected (high risk: 0.75). The required workflow reads a user-provided Lean/Business Model Canvas file, extracts its section text (e.g., Problem/Customer Segments/UVP/Solution/Revenue Streams) and then writes it into portfolio.json as canvas_context, which is subsequently used as LLM-visible context for creating downstream messaging—so the LLM is exposed to outsider-authored free text whenever the canvas is not authored by the operating user.