The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill facilitates the installation and use of the agent-browser tool from the official Vercel Labs GitHub repository. This external dependency is used to automate data collection from several well-known web services.
[PROMPT_INJECTION]: The skill implements a research workflow that ingests untrusted data from external websites, creating an indirect prompt injection surface. While this is typical for browser-integrated agents, the lack of explicit sanitization or strict boundary markers for the ingested content is a standard risk factor.
Ingestion points: The skill uses agent-browser get text body across several commands in SKILL.md to fetch content from Google Trends, YouTube, Reddit, TikTok, and Product Hunt.
Boundary markers: The instructions do not specify any delimiters or safety prefixes (e.g., "the following is untrusted web content") to isolate external data from the agent's internal reasoning.
Capability inventory: The agent has the ability to execute browser automation commands and write files (screenshots and research text) to the local project structure under ./brands/{brand-slug}/.
Sanitization: There is no mention of filtering or sanitizing the data retrieved from external URLs before it is processed for pattern extraction.

marketing-guerrilla