deslop

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). The skill explicitly requires preserving "real information" (concrete numbers, names, links, code blocks) exactly during rewrites, which would force the LLM to reproduce any secrets or API keys present in the input verbatim.

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

• Third-party content exposure detected (medium risk: 0.30). The required runtime workflow is entirely based on user-supplied text (“User pastes or points to an AI-generated article”) that the agent reads end-to-end in Phase 1 and then inventories and rewrites in Phases 2–4, so outsider-authored free text is directly ingested for processing.