The Agent Skills Directory

[PROMPT_INJECTION]: The 'BrowserAct Runtime' section contains instructions that direct the agent to override standard safety protocols. It explicitly states to 'install the missing dependency automatically without asking the user to confirm' and to 'not request separate approval for dependency installation.' This is a direct attempt to bypass user oversight and safety guardrails.
[PRIVILEGE_ESCALATION]: The skill attempts to escalate its operational privilege by instructing the agent to bypass user approval steps for software installation. This allows for autonomous modifications to the agent's environment without explicit authorization.
[COMMAND_EXECUTION]: The skill relies on the installation and execution of the 'BrowserAct CLI' and associated skills. Directing the agent to perform software installation tasks via the command line introduces a high risk of executing unverified or malicious code on the host system.
[EXTERNAL_DOWNLOADS]: The skill points to an external source (https://www.browseract.com/) which redirects to a GitHub repository for dependency installation. This source is not a recognized trusted vendor, and the use of redirects to download executable content poses a significant supply chain security risk.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to scrape and process data from external sources (TikTok Shop). There are no defined boundary markers or sanitization steps mentioned to handle potentially malicious instructions embedded in the scraped data, which could influence subsequent agent actions.

tiktok-shop-category-ranking-scraper