The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill instructions advise the user to install and run the chrome-devtools-mcp package using npx -y chrome-devtools-mcp@latest. This involves fetching code from the public NPM registry at runtime.
[REMOTE_CODE_EXECUTION]: By recommending the use of npx for the MCP server, the skill facilitates the execution of external code on the host system. This is a functional requirement for the skill's performance auditing capabilities but relies on the security of the third-party package.
[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it ingests and processes untrusted data from external websites (DOM, network requests, and performance traces) without explicit boundary markers or sanitization instructions.
Ingestion points: External URLs accessed via navigate_page, network headers retrieved through list_network_requests, and page snapshots taken via take_snapshot in SKILL.md.
Capability inventory: The skill uses tools to record performance traces, analyze network dependencies, and inspect the codebase.
Boundary markers: None; there are no instructions to the agent to treat external content as untrusted data or to ignore embedded instructions.
Sanitization: No evidence of sanitization or filtering of the retrieved content before it is processed by the agent.

web-perf