contentful-personalization

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

• Third-party content exposure detected (high risk: 0.80). The skill's workflow and reference code explicitly fetch and ingest Contentful entries (e.g., contentfulClient.getEntries / getAllExperiences, getAllAudiences in references/contentful-integration-guide.md and rendering-pipeline.md) and the live-debug sub-skill inspects arbitrary live URLs, meaning the agent will read and act on untrusted, user-authored third-party content from Contentful and arbitrary web pages as part of its runtime decisions.