paw-pa-pricing

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

• Third-party content exposure detected (medium risk: 0.65). Yes—runtime ingestion of outsider-authored free text can occur when the workflow reads {run-folder}/brief.md (structured brief content chosen by the user but not authored by the system) and especially {run-folder}/research-dossier.html/.md (explicitly treated as optional “client intel/pricingBenchmarks” from an external dossier), then generates pricing.json via mode references loaded into the LLM context.