The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill interacts with external content from trustradius.com, creating a surface for potential indirect prompt injection attacks.
Ingestion points: The trustradius_com__get_search, trustradius_com__read_page, and unbrowse.scrape tools fetch data from a public website.
Capability inventory: The skill includes tools to read, scrape, and perform browser-like actions (unbrowse.browse.act) based on site content. This includes an autofill feature for login actions.
Boundary markers: The instructions do not explicitly mandate strict delimiters or specific ignore-instructions for the external content.
Sanitization: Results are processed server-side by the vendor's service before being returned to the agent.
[EXTERNAL_DOWNLOADS]: The skill facilitates connection to a remote Model Context Protocol (MCP) server hosted by the vendor.
The server.json and installation instructions reference https://unbrowse.ai/mcp/trustradius.com.
As these resources belong to the official vendor domain, this represents standard functionality for a remote-hosted toolset.
[COMMAND_EXECUTION]: The README provides setup commands using npx and claude mcp add.
These commands are intended for initial configuration and point to the vendor's verified infrastructure.

trustradius