startup-launch-kit

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

• Third-party content exposure detected (high risk: 0.75). The orchestrator’s runtime LLM context can include outsider-authored free text when it invokes startup-audit in existing-project mode, because startup-audit reads the user-supplied codebase (e.g., README/docs/comments) and turns it into readable prose for the model; that codebase content is not authored by the operating user and is therefore outsider text fed into the agent context.