The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill facilitates the download of the chrome-devtools-mcp package from the npm registry via the npx command.
[REMOTE_CODE_EXECUTION]: By using npx, the skill executes the latest version of the chrome-devtools-mcp package directly from a remote repository. This package is managed by the official Chrome DevTools organization.
[COMMAND_EXECUTION]: The server allows the agent to perform actions such as taking screenshots and saving performance traces to local file paths, as well as managing browser instances.
[DYNAMIC_EXECUTION]: The evaluate tool provides the agent with the ability to execute arbitrary JavaScript code within the context of the browser, which is a powerful capability for scraping and interaction.
[INDIRECT_PROMPT_INJECTION]: The skill exposes the agent to content from external websites which can be ingested via navigation, network logs, and DOM evaluation. This creates a surface for indirect prompt injection if a malicious site attempts to exploit the agent's browser-control capabilities. 1. Ingestion points: Web page content accessed through navigate, network-logs, and the evaluate tool. 2. Boundary markers: No specific delimiters or instructions to ignore embedded commands are present in the provided interaction patterns. 3. Capability inventory: Includes arbitrary JavaScript execution, file system writes (screenshot), and comprehensive browser control. 4. Sanitization: There is no evidence of sanitization or content validation for data retrieved from the browser.

chrome-devtools-mcp-automation