The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill uses npx to execute the impeccable package. If the package is not present locally, npx will download it from the public npm registry. This introduces a dependency on external code that is not verified as a trusted vendor resource.
[REMOTE_CODE_EXECUTION]: The use of npx impeccable constitutes remote code execution, as it fetches and runs code from an external repository at runtime. Executing unpinned third-party packages presents a risk of executing malicious code if the registry or package is compromised.
[COMMAND_EXECUTION]: The skill requests and utilizes shell access through the Bash tool to run detection commands and start a background server (npx impeccable live &). Background processes can be used to maintain persistence or perform hidden actions during the session.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection. 1. Ingestion points: Source files (HTML, CSS, JS/TS) and browser console logs are read and processed by the agent. 2. Boundary markers: The instructions lack explicit delimiters or warnings to ignore embedded instructions in the ingested data. 3. Capability inventory: The agent has access to shell execution, browser script injection, and sub-agent spawning. 4. Sanitization: There is no evidence of sanitization or validation of the external content before it is processed by the model or its sub-agents.

critique