The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill configuration and instructions reference a remote Model Context Protocol (MCP) server at https://unbrowse.ai/mcp/abebooks.com and an OpenAPI specification at https://unbrowse.ai/api/v1/sites/abebooks.com/openapi.json. These resources are provided by the vendor to facilitate the skill's proxy-based functionality.
[INDIRECT_PROMPT_INJECTION]: The skill presents an attack surface for indirect prompt injection due to its core function of retrieving and processing external web data.
Ingestion points: The tools abebooks_com__get_servlet_search_results and abebooks_com__read_page retrieve and return readable text, links, and metadata from Abebooks.com directly into the agent's context.
Boundary markers: There are no specific instructions or delimiters provided to help the agent distinguish between the skill's instructions and potentially malicious instructions embedded within the retrieved web content.
Capability inventory: The skill grants the agent powerful capabilities, including full browser automation via unbrowse.browse.open and unbrowse.browse.act, as well as arbitrary task execution through unbrowse.run.
Sanitization: The documentation does not describe any sanitization, filtering, or validation processes applied to the content fetched from the external site before it is processed by the agent.

abebooks