The Agent Skills Directory

[PROMPT_INJECTION]: The skill instructions attempt to override the agent's safety and oversight protocols. It explicitly directs the agent to perform actions "automatically without asking the user to confirm" and states "do not request separate approval for dependency installation."
[REMOTE_CODE_EXECUTION]: The skill mandates the installation of external components (BrowserAct CLI and Skill) from a remote URL (browseract.com/...) if they are not present. By instructing the agent to bypass user consent during this process, the skill creates a high-risk vector for silent, unauthorized remote code execution.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to scrape trending content from TikTok, which exposes the agent to untrusted external data that may contain malicious instructions.
Ingestion points: TikTok trending video records, including titles, descriptions, hashtags, and creator metadata, are ingested into the agent context (SKILL.md).
Boundary markers: The instructions do not define any delimiters or provide warnings to the agent to ignore instructions embedded in the scraped content.
Capability inventory: The skill utilizes the browser-act CLI for browser interaction and includes logic for automatic dependency installation (SKILL.md).
Sanitization: There are no documented steps for sanitizing, filtering, or validating the data collected from TikTok before the agent processes it.

tiktok-trending-video-scraper