The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill instructions and usage patterns reference fetching the chrome-devtools-mcp package via npx -y chrome-devtools-mcp@latest. This downloads and executes the package from the official NPM registry.
[COMMAND_EXECUTION]: The skill uses the uxc link command to register persistent shell command aliases (such as chrome-devtools-mcp-cli) that execute the MCP server. It also uses command -v to check for existing local binaries.
[DYNAMIC_EXECUTION]: The skill provides an evaluate_script operation which allows the agent to execute arbitrary JavaScript within the context of the active browser page, as documented in references/usage-patterns.md.
[INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection by ingesting and processing untrusted content from the web.
Ingestion points: The skill reads external data via new_page, take_snapshot, list_network_requests, and list_console_messages (documented in SKILL.md and references/usage-patterns.md).
Boundary markers: The instructions do not define explicit delimiters or warnings to treat ingested web content as untrusted data.
Capability inventory: The skill can execute shell commands via the linked CLI, perform file uploads via upload_file, and execute JavaScript via evaluate_script.
Sanitization: No sanitization, filtering, or escaping of web-sourced content is specified before the data is processed by the agent.

chrome-devtools-mcp-skill