The Agent Skills Directory

[SAFE]: The skill's behavior is consistent with its stated purpose of developing pricing models and performing competitive research. It does not exhibit malicious patterns such as persistence, privilege escalation, or unauthorized data access.
[EXTERNAL_DOWNLOADS]: Fetches and installs the agent-browser skill from Vercel Labs' official GitHub repository. This is a trusted source used for browser automation.
[COMMAND_EXECUTION]: Utilizes the agent-browser CLI to perform web scraping and capture screenshots. These operations are essential for its research functions and are clearly defined in the instructions.
[PROMPT_INJECTION]: The skill processes untrusted data from external websites, creating a surface for indirect prompt injection. Ingestion points: Competitor pricing pages, G2 reviews, and Reddit posts accessed via agent-browser in SKILL.md. Boundary markers: None present to delimit external content. Capability inventory: File system writes to local brand directories and browser automation capabilities. Sanitization: No explicit filtering or validation of external content is specified.

marketing-pricing