The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to perform browser automation and web scraping, which naturally involves processing untrusted data that could contain malicious instructions.
Ingestion points: Data enters the agent's context from external websites via tools like Playwright and Puppeteer as described in the capabilities section.
Boundary markers: The instructions do not define boundary markers (such as XML tags or clear delimiters) or provide instructions for the agent to ignore embedded commands within the scraped data.
Capability inventory: The skill utilizes Playwright and Puppeteer, which have extensive capabilities including network access and potential file system interaction depending on the script execution environment.
Sanitization: The skill lacks any guidance or requirement for sanitizing, escaping, or validating the content retrieved from external sources before it is processed by the agent.

browser-automation