The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to use the lsof -i :<port> shell command to identify which ports are being used by local development servers. This is a legitimate use case for a development-oriented agent to ensure it is critiquing the correct running application. Evidence is found in references/task-walkthrough.md under the 'Find the Right URL First' section.
[INDIRECT_PROMPT_INJECTION]: The skill has a surface for indirect prompt injection as it is designed to ingest and process untrusted external data such as live web pages and application source code.
Ingestion points: Navigating to arbitrary URLs via agent-browser and reading local file content like USERS.md or component code.
Boundary markers: There are no explicit instructions or delimiters provided to the agent to treat external content as untrusted or to ignore instructions embedded within the analyzed data.
Capability inventory: The skill utilizes tools including agent-browser for interaction, chrome-devtools for inspection, and shell commands for environment discovery.
Sanitization: No specific sanitization or validation logic is defined for the ingested content before it is processed by the agent.

ux-critique