blog-series-drafting

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

• Third-party content exposure detected (high risk: 0.85). The skill’s mandatory workflow explicitly instructs the agent to “read every directly provided link first” and then follow internal links, meaning runtime-fetched public web content (outsider-authored pages) is ingested as readable text into the LLM context for source-grounded drafting.