The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The file references/installation.md instructs the user to globally install the chrome-devtools-mcp package using npm i chrome-devtools-mcp@latest -g. This downloads and installs code from the public npm registry at runtime.
[COMMAND_EXECUTION]: The skill operates by executing the chrome-devtools CLI tool to perform various browser automation tasks. This includes sensitive capabilities such as evaluate_script, which allows the agent to execute arbitrary JavaScript within the context of a web page.
[PROMPT_INJECTION]: The skill interacts with and processes data from external websites, creating a surface for Indirect Prompt Injection.
Ingestion points: The agent ingests untrusted data from web pages via take_snapshot (accessibility tree), list_console_messages, and list_network_requests as documented in SKILL.md and references/cli-reference.md.
Boundary markers: There are no instructions or delimiters provided to help the agent distinguish between its core instructions and potentially malicious instructions embedded in the web content it processes.
Capability inventory: The skill allows the agent to perform actions like clicking, filling forms, and executing JavaScript based on its interpretation of the ingested data.
Sanitization: No sanitization, validation, or filtering mechanisms are described for the content retrieved from the browser before the agent processes it.

chrome-devtools-cli