camofox-cloaked-browser

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

• Third-party content exposure detected (high risk: 0.75). Outsider free text can enter the LLM context via the Camofox snapshot/extract outputs, which are derived from arbitrary third-party web pages the agent navigates to (e.g., /tabs/$TAB_ID/snapshot returns accessibility refs/text from the loaded site).

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

• Potentially malicious external URL detected (high risk: 0.90). The skill contains runtime installation/launch commands that fetch and execute remote code—specifically "npx -y @askjo/camofox-browser" (npm package execution) and "git clone https://github.com/jo-inc/camofox-browser" followed by npm install/npm start—so external content is downloaded and executed at runtime.