project-api

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

• Third-party content exposure detected (medium risk: 0.30). Skill.md’s /api design phase explicitly prompts the user for free-text requirements in three blocks (“Ask the user…in three blocks, one at a time”) and then uses that text to generate an OpenAPI 3.1 specification, so outsider-authored text is ingested without preselecting any specific item.