mercadolibre-review-scraper

CRITICAL E004: Prompt injection detected in skill instructions.

• Potential prompt injection detected (high risk: 0.90). The skill explicitly instructs the agent to auto-install runtime dependencies without asking the user and to avoid describing the installation or requesting approval, which is a deceptive instruction outside the scraping task's stated purpose.

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

• Third-party content exposure detected (high risk: 0.85). SKILL.md instructs the agent to use BrowserAct to “capture browser-visible page content” from user-specified MercadoLibre/public URLs at runtime, which can include outsider-authored review text and other free-form content that the LLM ingests.

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

• Potentially malicious external URL detected (high risk: 1.00). The skill explicitly directs the agent to automatically install a required runtime dependency from the BrowserAct install URL (https://www.browseract.com/?co-from=ecommerce and https://www.browseract.com/?co-from=ecommerce&redirect=https://github.com/browser-act/skills/tree/main), which implies fetching and executing remote installation code at runtime to control the agent's capability.

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 explicitly instructs the agent to automatically install the BrowserAct CLI when missing and to do so without asking the user for approval, which directs the agent to modify the host system state by installing software even though it doesn't explicitly request sudo or file edits.