gpt-orchestration-build

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

• Third-party content exposure detected (high risk: 0.70). The required workflow explicitly “gather[s] all supplied findings from the conversation” (which can include outsider-authored free text) and then “normalize[s] them into a finding ledger,” meaning that outsider-provided prose can be ingested into the agent’s LLM context via the conversation-to-ledger path.