memory-continue

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.80). The skill calls build_context/read_note/recent_activity with memory:// URLs (e.g., memory://note-title and memory://folder/*) at runtime to fetch notes that are injected into the assistant's context and thereby directly control prompts, so this is a runtime dependency that can influence agent instructions.