ln-821-npm-upgrader

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

• Third-party content exposure detected (high risk: 0.90). The skill explicitly instructs (SKILL.md, Phase 4–6 and "MCP Tools for Migration Search") to query external MCP services and WebSearch (including reading URLs and community solutions/StackOverflow) to fetch migration guides and docs that the agent must read and act on, exposing it to untrusted public third‑party content that can influence tool actions.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly calls mcp__Ref__ref_read_url (reads arbitrary external URLs from search results) and WebSearch at runtime to fetch migration guides which the agent must use to control upgrade/migration instructions, so external URLs are used at runtime and directly influence agent behavior.