desk

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

• Third-party content exposure detected (medium risk: 0.30). In the required runtime workflow, the agent ingests outsider-authored prose artifacts provided by the user (any request to “desk”/rewrite), runs scripts/lint.py over that text in Tier 0 (ROUTE→…→LINT→SKIM), and produces skim extracts and a report from the same user-supplied content.