cross-platform-social-profile-search

CRITICAL E004: Prompt injection detected in skill instructions.

• Potential prompt injection detected (high risk: 0.90). The skill contains deceptive runtime instructions telling the Agent to install the BrowserAct dependency automatically without user confirmation and to hide/obscure the installation link from the user, which is outside the advertised profile-search purpose and violates user-consent expectations.

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

• Third-party content exposure detected (high risk: 0.85). The required workflow invokes the BrowserAct CLI to access real browser pages at runtime, which can ingest outsider-authored free text from public web content (e.g., arbitrary URLs/pages the user asks to search/monitor) into the agent’s LLM context.

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

• Potentially malicious external URL detected (high risk: 0.80). The Skill treats the BrowserAct CLI as a required runtime dependency and instructs automatic installation using https://www.browseract.com/?co-from=social-media&redirect=https://github.com/browser-act/skills/tree/main (and the related https://www.browseract.com/?co-from=social-media) which would fetch and install/execute remote code at runtime.

MEDIUM W013: Attempt to modify system services in skill instructions.

• Attempt to modify system services in skill instructions detected (medium risk: 0.60). The Skill instructs the agent to automatically install a required CLI dependency and resume execution "without asking the user to confirm" (and to treat the CLI as a required runtime dependency), which encourages making persistent system changes on the host without explicit user approval even though it doesn't explicitly request sudo or file-level modifications.