polars

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

• Third-party content exposure detected (high risk: 0.80). The skill includes explicit examples that fetch data from arbitrary remote URLs (e.g., pl.read_csv("https://example.com/data.csv") and various pl.scan_/pl.read_ calls that accept user-provided URLs/paths), so the agent could ingest untrusted public content as part of its workflow.