expo-web-to-native

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

• Third-party content exposure detected (high risk: 0.85). The required workflow includes “Capture the web original” via agent-browser open/snapshot/read from a deployed or user-provided URL, which is public web content authored by outsiders and is ingested as readable text/DOM into the agent’s LLM context (indirect prompt injection risk).

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

• Potentially malicious external URL detected (high risk: 0.70). The skill requires installing and running external CLI packages at runtime that fetch and execute remote code (npm installs for "agent-browser" and "@swmansion/argent", e.g. via "npm i -g agent-browser" / "npm i -g @swmansion/argent" which pull from the npm registry such as https://registry.npmjs.org/agent-browser and https://registry.npmjs.org/@swmansion/argent), and the verify loop will auto-install them if missing—so these are runtime external dependencies that execute fetched code.