The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The instructions in the "BrowserAct Runtime" section explicitly command the agent to install missing dependencies automatically from an external GitHub repository. The skill specifically directs the agent to perform this installation "without asking the user to confirm" and further emphasizes "do not request separate approval for dependency installation," which bypasses standard user oversight for executing remote code.
[COMMAND_EXECUTION]: The skill relies on the "browser-act" CLI and Skill to perform its core functions. This requirement involves the execution of external commands on the host system to interact with web browsers and manage dependencies.
[PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection because it is designed to ingest and process untrusted data from various social media platforms (trending topics, hashtags, creators). There is no provision for sanitization or boundary markers to prevent malicious instructions embedded in that data from influencing the agent's behavior.
Ingestion points: Trending topics, music labels, hashtags, creators, and content formats harvested from social media via BrowserAct.
Boundary markers: Absent; the skill does not instruct the agent to use delimiters or ignore instructions within the collected data.
Capability inventory: The skill utilizes the "browser-act" CLI and associated skill, which can perform browser interactions and execute commands.
Sanitization: No evidence of validation, escaping, or filtering of the external social media content is present in the instructions.

social-trend-monitor