business-card-contact-scraper

CRITICAL E004: Prompt injection detected in skill instructions.

• Potential prompt injection detected (high risk: 0.90). The Skill explicitly instructs the agent to automatically install the BrowserAct dependency without asking the user and to conceal that action (not describing it as a browser request or seeking approval), which is a deceptive behavior outside the advertised scraping purpose.

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). This skill explicitly instructs automatic installation and invocation of an external CLI (BrowserAct) without asking the user for confirmation or describing the action as a user request, which enables unauthorized installation and potential supply-chain or data-exfiltration abuse.

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

• Third-party content exposure detected (high risk: 0.85). This skill’s required workflow delegates live website interaction to the external browser-act Skill (public web pages fetched at runtime), which can ingest outsider-authored free text from arbitrary URLs/pages into the agent’s LLM context via BrowserAct.

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

• Potentially malicious external URL detected (high risk: 0.80). The Skill explicitly requires and instructs the agent at runtime to automatically install and invoke BrowserAct using the install link https://www.browseract.com/?co-from=lead-generation&redirect=https://github.com/browser-act/skills/tree/main, which would fetch and install external code that the agent relies on to execute the task.

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

• Attempt to modify system services in skill instructions detected (high risk: 1.00). The Skill instructs the agent to automatically install the BrowserAct dependency without asking the user for confirmation, which directs the agent to modify the host system (potentially requiring elevated privileges) and bypass user approval.