The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes untrusted content from external websites. \n
Ingestion points: The browser_extract and browser_navigate tools (defined in SKILL.md) allow the agent to fetch and read content from any URL. \n
Boundary markers: There are no explicit delimiters or instructions to the agent to ignore potentially malicious instructions embedded within the scraped content. \n
Capability inventory: The agent has access to powerful tools such as Bash and Write as specified in the allowed-tools frontmatter. \n
Sanitization: No content validation or sanitization is mentioned for data retrieved from the browser. \n- [EXTERNAL_DOWNLOADS]: The skill setup involves running the browser-use package via uvx, which is a standard procedure for using this browser automation tool.

browser-automation