mise

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

• Third-party content exposure detected (high risk: 1.00). The skill documentation explicitly describes fetching and executing remote, user-provided content (e.g., references/tasks.md "Remote Tasks" with file = "https://example.com/build.sh" and git::https://github.com/...", plus backends like http/github and hooks in references/advanced.md), so the agent could ingest untrusted third‑party content that can change behavior or drive actions.