javascript-sdk

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

• Third-party content exposure detected (high risk: 0.90). The skill's SKILL.md and referenced files explicitly show agents/tools that perform web searches and browser automation (e.g., the RAG pattern using appTool('search','tavily/search-assistant@latest') in references/agent-patterns.md, internalTools().webSearch(true) in references/tool-builder.md, and browser-automation examples in references/sessions.md) and also accepts remote URLs/files (references/files.md), which means the agent is instructed to fetch and interpret untrusted public web content that can influence its actions.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill examples include a skills entry with url "https://example.com/skills/api-docs.md", which is presented as a skill that would be fetched and injected into an agent's context at runtime, so remote content could directly control prompts/instructions.