ai-slop-document-auditor

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

• Third-party content exposure detected (high risk: 0.75). The skill ingests user-provided documents (including arbitrary files/URLs) and extracts their readable text into document_units, which is then passed into subagents via CONTEXT: <document_units>—so if the input document text is authored by an outsider, that outsider free text becomes LLM context through the extraction→document_units→detector/guardrail delegation path.