The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill utilizes npx chrome-devtools-mcp@latest to download and execute the Model Context Protocol (MCP) server from the npm registry.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and interact with untrusted data from external web pages, creating a vulnerability surface for indirect prompt injection.
Ingestion points: Untrusted data enters the agent's context through take_snapshot, evaluate_script, and get_css_styles after navigating to a URL.
Boundary markers: No explicit delimiters or instructions are provided to the agent to distinguish between its system instructions and the content retrieved from web pages.
Capability inventory: The skill possesses significant capabilities that could be abused if an injection is successful, including arbitrary script execution via evaluate_script, form filling via fill, and the installation of browser extensions via install_extension.
Sanitization: There is no evidence of sanitization or filtering of the retrieved DOM content or script results before they are processed by the agent.

chrome-devtools